Alex Urosevic Importance of Demand Forecasting within two case companies focusing on production and warehousing Masters Thesis Vaasa 2024 School of Technology and Innovations Master’s thesis in Industrial Management Programme 2 UNIVERSITY OF VAASA School of Technology and Innovations Author: Alex Urosevic Title of the Thesis: Importance of Demand Forecasting within two case com- panies focusing on production and warehousing Degree: Master’s Program in Industrial Management Programme: Industrial Management Supervisor: Ahm Shamsuzzoha Year: 2024 Page Number: 84 ABSTRACT : The objective of this research is to explore the importance of demand forecasting within pro- duction and warehousing in the coffee industry, with a particular focus on identifying key varia- bles that impact decision making. The main motivation of starting this research came from real- izing how volatile and changing our markets are around the globe and how certain factors can trigger a domino effect on the supply chain. With the current supply chain processes being very volatile and prone to quick changes, it is highly important to see which factors have major impact on processes. This study examines the practices of case company 1 as well as 2, which share the same raw materials but operate in different production locations, however, serve in the same market. Through a qualitative approach and structured interviews, the research uncovers critical factors such as lead time, communication, as well as consumer purchasing trends that influence demand forecasting practices in warehousing as well as production. Key areas include the role of communication with stakeholders, adapting to market changes, as well as understanding the importance of marketing campaigns. The research revolves around three main research ques- tions: (1) what role technology play in enhancing demand forecasting processes in production and warehousing? (2) what are they key factors influencing forecasting accuracy, as well as (3) how does collaboration between stakeholders improve the forecasting accuracy in the coffee supply chain? The study findings show that communication with stakeholders working the sup- ply chain, as well as focusing on understanding campaign, and flexibility are crucial roles in fore- casting. understanding these dynamics to optimize inventory management, improve decision making, delivery and sales. KEYWORDS: Supply chain, warehousing, process, function, collaboration, demand forecasting 3 Table of Contents 1 Introduction 5 1.1 Research Background 5 1.2 The purpose of the study and research question 6 1.3 Methodological choices of the research 7 1.4 Study aims and research gaps with objectives 8 1.5 Study contributions 9 1.6 Research Structure 9 1.7 Justification of research questions 9 2 Theoretical framework 10 2.1 Different hypothesis approaches 11 2.2 Defining Forecasting 11 2.3 Forecasting models 12 2.3.1 Quantitative Forecast 13 2.3.2 Qualitative Forecast 14 2.3.3 Most Beneficial method for Demand Forecasting in the coffee supply chain 15 2.4 How technology is affecting demand forecasting with AI and machine learning 16 2.5 Inventory Management benefits of an Enterprise Resource Planning System 17 2.6 Unpredicted changes in the global market which affect forecasting 18 2.7 Demand forecasting within production 19 2.8 Demand forecasting within warehousing 19 2.8.1 Demand forecasting order management 20 2.8.2 Delivery Management 21 2.8.3 Customer Management 22 2.8.4 Sales Management 22 2.9 Comparison between production and warehousing demand forecasting 23 2.10 Tools to help forecasting within warehousing 24 2.10.1 Warehouse ABC Analysis 25 2.10.2 Batch Tracking and coffee traceability 27 2.10.3 Inventory turnover analysis 28 4 2.10.4 Third party logistics warehouses 29 2.11 Tools to help forecasting within production 30 2.11.1 Trend projection 31 2.11.2 Lifecycle analysis method 32 2.11.3 The moving average method 33 2.11.4 Time series analysis 34 2.12 Enterprise resource planning and manufacturing execution system 34 2.13 Demand forecasting SWOT analysis 35 2.14 Importance of silos and monitoring stock for coffee pre-production 36 3 Methodology and research approach 37 3.1 Case Study Companies 38 3.2 Data Collection 39 3.3 Discussion and recommendations 40 3.4 SWOT analysis around structured interviews 41 3.5 Analysis of sample data 42 3.6 Sample size calculation and analysis 43 3.7 Managerial implications 44 4 Study results and analysis 46 4.1 Study findings and recommendations for future managers 47 5 Conclusion 48 5.1 Key findings and limitations 49 5.2 Future research suggestions 50 6 References 51 7 Appendices 61 7.1 Research Questions 61 7.2 Sample Data 83 5 1 Introduction Demand forecasting is the process of using predictive analysis and historical data to pre- dict upcoming events in the near or far future. This process helps businesses prepare for peak seasons, sudden customer demand, and mitigate situations where they would have nothing to sell to end suppliers. From scheduling production to the end products stored in warehouses, understanding fluctuating markets is something that analysts are craving to understand more as the global market develops. With fully understanding demand forecasting, a team can develop a pricing strategy with much more ease as they under- stand exactly when prices may jump or fall. Throughout this master’s thesis, various methods and techniques will be addressed that can help a company or team get the most out of demand forecasting. Geography, competition, as well as seasonality are all very important factors to take into consideration when predicting upcoming events re- garding production line running as well as end warehousing (Hand, 2024). 1.1 Research Background The goal of this research was to seek and dig deeper into the importance of demand forecasting within production and end warehousing. Production as well as warehousing work closely together, as they are the two links within the supply chain process. With the use of demand forecasting tools and models, one can justify the importance of plan- ning ahead from two standpoints, raw materials into production, as well as then putting the product within the warehouse before it leaves for its dedicated customer. The con- cept of forecasting, as well as demand planning are studied throughout the literature with the help of the case study from a coffee production and warehousing standpoint used in the case study example. As seen in the past with the rapid pace of evolving mar- kets around the globe, it is vital for companies to understand raw material supply as well as production capability, all the way to the amount of space there is within a facility to store goods in a safe and appropriate manner. Various questions have been addressed by analysts within supply chain regarding demand and linking two parts of the supply chain, such as: 6 - How many units of inventory do we need in safety stock to have enough per SKU? - How often do we need to replenish stock? - How do we manage inventory if we do not know production scheduling well? - Where do we expect to be in a year from now? All these questions above show the importance of knowing the business and needs of not only the production line raw materials, but space within the warehouse as well as the demand per season, for example, from the end consumers. All in all, forecasting pro- jections is very hard as there are so many different variables that are in play, and with the use of qualitative data and different models, experts can predict and prepare for upcoming changes (Diezhandino, 2022). 1.2 The purpose of the study and research question The purpose of this study is to see how important it is to collaborate and understand production line raw material needs, as well as balance inventory stock accordingly throughout the years different peak seasons. Comparing different situations where pro- duction is low with lots of raw materials, versus times when production should be work- ing nonstop but raw material inbound is minimal. The case study presented in this paper will elaborate different situations within the case company presented, as well as cover different open questions with answers conducted through interviews done with the sub- ject matter aspects within the field. The following research questions will be focused on through this thesis - What role does technology play in enhancing demand forecasting processes in coffee production and warehousing? - What are the key factors influencing forecasting accuracy in the coffee industry? - How does collaboration between stakeholders improve the accuracy of de- mand forecasting in the coffee supply chain? With technology developing and market trends evolving rapidly, a clear view on the im- pact of artificial intelligence and machine learning on demand forecasting is vital to un- derstand. With the case study focusing on demand forecasting within production and warehousing within the coffee industry, raw material and its availability plays a key role 7 in understanding inventory levels and seasonal trends. Various data was taken from the case study to compare and find the correct answers to these research questions listed above. Various subject matter experts were interviewed, and the questions were ana- lyzed to make justified assumptions. 1.3 Methodological choices of the research In this research on the importance of demand forecasting within production and ware- housing in the coffee industry, a qualitative methodological approach has been chosen to provide a comprehensive understanding of the nuanced challenges and practices within this sector. The primary data collection will involve in-depth interviews with ten subject matter experts, allowing for rich, detailed insights into their experiences and per- spectives on demand forecasting. This qualitative approach facilitates an exploration of the complexities inherent in forecasting processes, including the influence of market trends, consumer behavior, and supply chain dynamics. By employing thematic analysis to identify key patterns and themes for the interviews, the research aims to uncover best practices and common obstacles faced by industry stakeholders. Additionally, a case study framework will be utilized to contextualize findings within specific companies or regions, enhancing the analysis. Using this type of methodological choice emphasizes and helps gain deeper insights into specific phenomena- in this case demand forecasting within the coffee industry. It allows and helps identify patterns, best practices, and challenges across different segments of the industry. This includes producers, distributors, as well as retailers. In addition to the interviews conducted, we will consider supplementary data, modified reports from the case company due to confidential issues, to provide a broader context for the findings found. We will conduct the comparative analysis by seeking forecasting methods, tech- nology utilization, as well as responses to market changes. 8 1.4 Study aims and research gaps with objectives The objective of this research is to determine how important demand forecasting is within production and warehousing, as well as whether there are certain variables that have a big impact on decisions being made. With technology evolving and different tools used for demand forecasting being enhanced annually, case company 1 as well as 2 must be careful and understand the evolving market. With the in-depth interviews focused on subject matter experts within the case companies, a clear view on what key factors take place when considering demand forecasting within production as well as warehousing. Both steps within the supply chain are demandant on materials and inventory, and there- fore we can easily make judgments on what drive decisions in certain situations when planning. Both case company 1 as well as 2 use the same raw materials, however, are produced in two different places. The end consumers are within the same market, and within this research we can seek different answers to how they can help each other in difficult situations to reach the end goal and provide ready coffee to end consumers. The aim of this research is to seek the importance of demand forecasting within produc- tion and warehousing, and at the same time avoid common pitfalls that may occur, to meet the study objectives listed above. It is highly important to be specific, clear, rele- vant, as well as realistic when writing research aims and objectives (Thomas, 2023). Eval- uating and seeking the importance of certain variables that revolve around accurate de- mand forecasting, for example, inventory management, raw materials, and resources for production as well as warehousing. Interviewing those that work directly within the sup- ply chain around production as well as warehousing, this research paper aims to find the key factors involving successful supply chain processes. Potential research gaps revolve around the focus on climate change, and how it may potentially fully impact the supply chain and coffee supply in the near or far future. Currently, forecasting is being done on previous data, and not in the future, experts cannot know what kind of effect global warming can have on coffee supply in coming years. Coffee demand being very seasonal, it has many influential factors on harvest cycles and global events. Traditional forecasting methods may struggle to keep up to account for such fluctuations that may occur. 9 1.5 Study contributions Within the context of the coffee industry and supply chain, specifically within production and warehousing, accurate demand forecasting enables businesses to prepare for times annually when sales tend to exponentially increase. Various study contributions within this field include optimizing inventory management within warehousing, as well as en- hancing production planning as well as scheduling. Throughout the research, one can see the importance of planning and optimization of warehouse space utilization, improv- ing lead time management, as well as using automation and technology to its full poten- tial. Flexible as well as adaptive supply chain models must be in use, which is presented in this research, as lead times for raw materials are longer than in many other industries. Collaboration in supply chain management can be divided into many main key factors for example information sharing, joint planning and forecasting, as well as coordinated operations (LastMileLogistics, 2024). 1.6 Research Structure This thesis is structured in a way where the literature review is opened into different main factors affecting demand forecasting within production and warehousing. The main ideas were taken from the theoretical framework and brainstormed to create the best possible questions for subject matter experts working within the field. A clear approach as to why the qualitative method was used and not a quantitative was also justified. A SWOT analysis was made around demand forecasting based on theoretical framework, as well as an analysis on the structured interviews that took place. Different methods of forecasting within production and warehousing were presented to give the reader a clear view on the main factors that must be taken into consideration throughout the supply chain. 1.7 Justification of research questions The following research questions below which were mentioned already previously have a huge impact on many other industrial supply chain segments as well. Within the coffee segment however, there are various reasons as to why it has a major effect on this one specifically. Coffee in Europe must be shipped across the globe and therefore tracking 10 from raw material to end products is never done without the use of technology and ac- curate communication. Without knowing estimate times of arrivals or possible changes in the production, it is almost impossible to have ready made coffee packages produced to customers on time. Key factors that influence forecasting accuracy within the supply chain are the different crucial components of communication, adapting to change, as well as understanding consumer behavior and always have enough safety stock. Without forecasting proper amounts, with coffee having best before dates and certain regulations for quality, it is very difficult to understand demand forecasting within the field without pointing at key influencing factors. In many supply chain areas of different industries, collaboration can in many cases be more laid back, whereas in the coffee industry, col- laboration with stakeholders is crucial as processes tend to change weekly. If communi- cation does not work within the different sectors of the supply chain of a company, as well as outsourced stakeholders, inaccurate sales and stock amounts can start popping up, leading to bad sales. 2 Theoretical framework Forecasting has been an ever so evolving topic throughout the past decade and tends to get more and more precise with the innovative new technology evolving. With the use of different tools and methods, one can make better assumptions with customers and end consumers and therefore prepare production as well as warehousing to its greatest potential. Seeking minimal risk as well as maximizing different utilities is always the end goal, however different factors come into play that may not make this so easy. All in all, it is critical for those making assumptions and decisions to understand possible market trends in the future using past data to optimize sales and processes to their full potential (Joiner, 2023). In the following chapters defining different models as well as the scope definition of forecasting will be addressed. Further we will dig deeper into the impact on artificial intelligence and what role it plays in modern day demand forecasting. Lastly, we will look at inventory management as well as the comparison between forecasting within 11 production as well as warehousing and how they play a key role within one another. There are so many different factors that come into play, and specialists tackling the com- petitive machine learning results, wanting to always verify and see that the assumptions that a computer module is creating are not incorrect. 2.1 Different hypothesis approaches Throughout this case study, the aim is to find different hypothesis approaches and see whether the outcome is the one that we initially were looking for. These include whether technology plays a key role in demand forecasting, as well as whether there are certain factors that must come into play for the supply chain to successfully execute. How in fact demand forecasting can improve production, whether it can mitigate complications within the supply chain, as well as how it affects business and customer satisfaction, all must be investigated. To fulfill the expectations and customer needs within end products, various methods within demand forecasting can be utilized such as passive, active, short- term, as well as long term forecasting methods. Using customer survey methods, expert opinion methods, as well as statistical methods can all help proactivity when seeking the best knowledge on amounts needed to be produced as well as stored in warehouses per different product category (Champion, 2024). 2.2 Defining Forecasting Planning within the supply chain can be difficult at times, as expected, and unexpected disruptions can develop at any given point in time. Accurate forecasting uses data from the past as well as insights into what is predicted to happen in the near or far future. Staying on top of inventory, resources within production and warehousing, budget plan- ning, as well as transportation booking are all vital points to investigate. Each procedure relies on one another as they happen one after another within the supply chain cycle. Strategic planning is one forecasting aspect which considers already known markets, however, takes into consideration possible expansions that may happen in the future. Data quality and quantity are vital for correct estimates to be made for the near or far 12 future which are all challenges companies face (Owczarek, 2023). Forecasting can be broken down into two different sectors, qualitative as well as quantitative. Quantitative forecasting relies on historical data to predict what will be needed. On the other hand, if previous data is harder to find, we can rely on qualitative forecasting as it seeks new insights as well as expertise within the field (Maersk, 2023). Complex patterns that AI can make assumptions out of, in the past as well as in current trends, minimizes the number of errors that a human may make when doing the same task manually. The capability of algorithms adapting to market fluctuations marks a great advantage when considering the extremely volatile environment we live in. As we move forward with AI advancing and making complex models, it is making specialists harder to extract what these models tell us exactly. Misjudged AI decisions can on the other hand lead to huge mistakes in demand forecasting if one does not know how to read the mod- els created (Buehler, 2024). 2.3 Forecasting models As mentioned in the previous chapter, having a clear distinguish between the qualitative and quantitative forecasting models, and which one is being used, can easily be broken down into sub sections within either one of the two. Both methods have their positive and negative aspects and can easily be combined to dive deeper into analyzing trends and ways of executing decisions. The following figure below shows both methods and a clear view on the good and bad sides of both forecasting methods. As we can see there are various pros and cons for both the quantitative and qualitative forecasting methods. The study being around a food and beverage sector, using a qualitative approach allows great interpretation and comprehensive understanding of what expectations are to be met, as well as plays a key role in predicting consumer behavior annually. 13 Table 1: Positive/Negative approaches to quantitative as well as qualitative forecast models (IndeedEditorialTeam, 2024) The two forecasting models can be successfully used when considering looking at de- mand needed in the future, however, can get complicated as making coffee depends on the availability of green coffee around the globe. The correct inventory amount needed in production of green coffee must be forecasted correctly as many containers that are shipped may be claims and have coffee that does not pass quality standards (Akarsu, 2024). 2.3.1 Quantitative Forecast Within quantitative forecasting, various models such as moving average, exponential smoothing, as well as ARIMA can be used with historical data to identify patterns and project them in the future. Moving average in simple terms is used to calculate and iden- tify the direction of a trend with taken subsets rather than considering every simple data point. Specialists commonly think of this forecasting model as averaging and can be used for a single time or many broken down periods (Pollock, 2022). Exponential smoothing methods on the other hand use a variety of different weighted averages from the past to predict future trends. These combine error, trend, as well as seasonal components in 14 a smoothing calculation. In other terms, these are referred to as ETS. They do not need to be merged, and all used in the calculation, however, can be (Franco, 2022). Arima modeling is a robust tool for time series forecasting, which combines both past errors and values to predict future values. This modeling technique is used in sales, stock, as well as predicting prices. Within the model, there are three different variables ex- pressed. P, D, as well as Q which stand for number of lag observations, number of times the data needs to be differed, as well as the number of lagged forecast errors (Hayes, 2024). This works by identifying the values of P, D, as well as Q by analyzing the data (using plots and statistical sets). Next steps are to estimate by fitting the model to his- torical data, diagnosing to check the accuracy as well as adjust if needed, and lastly im- planting this and forecasting various values. The Arima model can be well used in fore- casting coffee by collecting sales data from for example the past 5 years, preprocess it by removing seasonality and trends, identifying ARIMA parameters, and lastly fitting it in the model and generating an annual sales forecast. This is way of working is great for analyzing time series data to better dive deeper into understanding previous processes as well as forecasting future values needed for production (Noble, 2024). 2.3.2 Qualitative Forecast The qualitative demand forecasting method differs quite drastically from the quantita- tive approach as it relies on subjective judgment, expertise, as well as intuition rather than quantitative data. Various methods can be used within this forecasting method, such as the Delphi method, market research, expert judgement, scenario writing, as well as sales force composite. All these different methods above rely on human judgment and base off insights rather than numerical data. The accuracy of data depends a lot on the experience of the specialists involved (Chapman, 2024). The Delphi model brings a variety of great ideas to use as it uses expert panels, keeps questions anonymous, as well as includes multiple rounds of questions to dive deeper into what is trying to be asked, and provides clearer answers every time. This process 15 also gives great feedback and adds revision to see how processes can be done differently in the future. To successfully conduct the Delphi model, there are a few steps that need to be followed. Firstly, defining the problem, which after experts are selected, which af- ter the first round of questions can be gone through (Indeed, 2023). There are various positive and negative sides to using the Delphi model. Gaining knowledge from experts, receiving honest answers, as well as reaching a strong group consensus are all great rewards of utilizing this method. However, this method prevents live discussion and may be slower than other methods as it requires several rounds of questions, and some answers provided may not lead to any good judgment of the ques- tion or task asked (Bryne, 2024). 2.3.3 Most Beneficial method for Demand Forecasting in the coffee supply chain Demand forecasting in the coffee industry can be difficult at times due to the variety of factors that must be looked at. Consumption, weather conditions, seasonal trends, and consumer preferences all greatly impact coffee for production and end warehousing. Correct data collection and preprocessing, model selection and training, as well as pre- cise visualization and communication are all vital steps when looking to get the correct outcomes (Greene, 2024). Machine learning is innovative technology that enhances sub- ject matter experts to dive deeper into different variants to further understand the con- sumption and demand for coffee. Tim series forecasting to forecast specific future trends of sales based on cycles, as well as the use of different linear models can both find very precise predictions in the near and far future. Various holdbacks can occur when trying to use machine learning, as the correct input of data must be applied to get a proper outcome. Data quality and integration, the dif- ferent dynamic market conditions, as well as scalability can all set constraints on getting the correct information an expert wants to analyze. Overall, there is no specific method which is best used for demand forecasting coffee, but rather utilizing different methods 16 within machine learning can demonstrate potential outcomes to enhance and predict what the future holds (Akarsu, 2024). 2.4 How technology is affecting demand forecasting with AI and machine learning The leaning step in artificial intelligence and utilizing mass amounts of data in a matter of seconds to make assumptions has been a vital change in demand forecasting ever since it was first introduced. Humans can do all of what artificial intelligence is already doing, however timing plays a key role when making decisions with raw materials and what needs to be produced. Extensive manual input that humans do when creating de- mand forecasting assumptions is only a minor amount of data utilized. According to researchers, 10% of previous data is used when making forecasting models, whereas closer to 100% of data is used by AI. The capability to merge data from various sources and find common trends is a huge step forward in forecasting as it saves time and re- sources (Pacemaker, 2024). To a certain extent, experts still need to verify whether these machine learning models used with artificial intelligence are valid, and not only rely on the output they give. A notable example was a huge disruption caused by the pandemic in 2020, as there was such a huge, unprecedented impact on consumer behavior and market trends (Dunlea, 2024). Warehouse space optimization plays a key role in saving costs as well as storing goods in a safe and appropriate manner. AI driven demand forecasting for warehousing space an- alyzes SKU-level data, order history, as well as looks at market trends to find the best fit for storing goods in a certain warehouse. This helps predict which products are more likely to move faster, as well as helps the workers within the warehouse save time and not have to shuffle pallets or stored goods from one place to another. This can be well used with replenishment with artificial intelligence. AI can predict seasonal trends and sales and let the warehouse managers know where to best put goods that tend to move quickly (LeewayHertz, 2024). 17 2.5 Inventory Management benefits of an Enterprise Resource Planning System Inventory management is critically important in supply chain management as it ensures the correct balance between stock available and demand, helps optimize cost, as well as prevents disruptions from not being able to provide goods to end consumers. With the huge amounts of production using manufacturing enterprise systems, otherwise known short for MES systems, companies can track consumption and material flow with great precision. Building on top of MES, which is used in production, using an enterprise re- sourcing planning system helps inventory management as the finished material flow is stored and easily reached. An ERP system integrates core retail business processes and bundles all parts of supply chain operations. Inventory being one of retail businesses' key assets, it is crucial to be able to track and use tools to manage these resources. ERP systems combine purchasing, sales, finance, as well as all other supply chain related op- tions into one whole, giving a company the capability to trace each step of what is being done (Field, 2023). Inventory management within an ERP can be divided into five differ- ent segments as shown in the figure below. Each and every stage involving real-time inventory tracking, reporting and analytics, integration with sales and purchasing, de- mand forecasting, as well as automated reordering are vital when using a ERP software. 18 Figure 1: ERP inventory management key features (Field, 2023) Real time inventory tracking is vital, especially in retail and the coffee industry. Best be- fore dates as well as expiry dates are tracked at a master data level within the ERP, and therefore knowing what stock is within a warehouse is crucial. Visibility into what needs to move quicker, as well as what stock needs to be replenished are key factors that an ERP can be used for. Efficiency as well as cost savings, a feature that optimize stock man- agement by reducing the need for various physical inventory counts needed which re- quire resources by employees within a company. Automated recording, integration within sales and purchasing, as well as reporting and analytics can all help within inven- tory management within an ERP system (Raj, 2024). 2.6 Unpredicted changes in the global market which affect forecasting There are various factors that highlight the importance of being prepared for unpre- dicted changes in the global market. Impact on production capability, warehousing, ge- opolitical instability, as well as other changes such as consumer trends all have a huge 19 effect on changing environments. Many consumers and even subject matter experts ap- proach unpredicted changes in the global market from a negative standpoint, however there are many different positive changes that can occur and affect the whole supply chain. Emerging markets, regional expansion, and even geographic growth can all impact production, warehousing and the coffee market. Demographic shifts have been seen in the past years, as for example consumers are starting to drink coffee at a younger age. Coffee has also become a trendy beverage and has been made popular amongst the younger generation. Whether it is a cold brew coffee or a traditional coffee, the main ingredient being the coffee bean is still used no matter what form the end drink is made of (GourmetPro, 2024). 2.7 Demand forecasting within production Accurate forecasting in manufacturing has a domino effect on many other factors if not performed well and with a structured approach. Understanding the maximum produc- tion pace capability as well as working along with the end warehouse to understand availability for storage are both vital when considering demand forecasting within pro- duction. Employee amounts for days of production that may not be so large, claim han- dling, bug errors in the manufacturing enterprise system being used, and any other situ- ations that may be brought up must be considered when forecasting within production. Demand forecasting for the next week's production cycle usually gone through a week before. Subject matter experts from logistics, integrated business planning, as well as production teams sit together to see what resources they have as well as what the pro- jected amounts are wanted to be made. All in all, efficient resource allocation with in- sights into the future demand, manufacturers can allocate resources to ensure smooth production operations (planettogether, 2024). 2.8 Demand forecasting within warehousing There are various factors that play a key role when demand forecasting within warehous- ing. Forecasting provides a framework for accessing detailed information about one or more warehouses' current inventory. Sales play a key role when considering inventory within a warehouse, as it is the key driver that helps push stock out of the warehouse to 20 end consumers. This especially has a key role when not only looking at short term, but also long-term sales. Relying on past data and trends to project future demand is what many companies tend to do, however, may not always be the best way to go as new scenarios evolve from year to year (Mecalux, 2024). Supply chain conditions as well as customer behavior all have a huge impact on demand forecasting within warehousing. Setting down the foundation to a proper guideline to demand forecasting must involve many other teams not just those that work within lo- gistics. Unusual demand outliers can be the result of known actions, for example sales promotions, or even large one-time orders (Borgman, 2022). A great example used from Case Company 1 has shown that all bulk products that are sent to specific customers are relocated from the production site which has a warehouse attached to it, to a 3PL part- ner to save storage and have more space for goods that are replenished in smaller stor- age locations. Goods that are sold the most and in large amounts are sent directly from the production line via trucks to an outsourced warehouse, where they are stored and shipped in bulk to two main customers that only buy in bulk. In this scenario, it is easier to handle the goods that are sold in smaller amounts in a separate warehouse. 2.8.1 Demand forecasting order management Order management plays a key role in accurate forecasting and reduction of stock outs. By analyzing past orders and sales, you can predict future demand more accurately. All these helps identify different patterns and trends. Customer service as well as sales must work hand in hand to communicate with one another for a smooth supply chain process to happen. Certain annual trends for example Christmas or easter campaigns are usually known up front ahead of time. The sales team gets a forecast on how much a certain customer will buy, after which the supply team can start looking at production amounts. During this same time the warehousing and logistics team have the rough amounts of what will need to be stored soon and can start preparing and planning. This all is scat- tered and effects accuracy, inventory amounts, cost management, as well as involves four different order management steps; receive order, fulfill order, manage inventory, as well as interact post-sales (Jenkins, 2022). 21 Within an ERP system there can be certain customers that order with an EDI integration, where the order comes in manually, however, there are larger customer deliveries that are then manually made by customer service workers within the company. Within order management, it is vital to understand the certain batch that a customer may want. If this is not possible to provide, clear communication must be undertaken for verification from the customer whether a certain other batch can be sent out. All these steps are linked with one another. If goods are sold from a 3PL warehouse from where the customer service team checks the stock, the stock must be up to date as that is in batches and amounts that are in the ERP are the amounts that they look at when making orders. This step shows the importance of a logistics team worker to check that integration and stock levels are correct, and if they are not, corrections should be made as soon as possible. In many cases, order management creation is done with batches that are in the ERP but not physically within the warehouse when a warehouse worker goes to collect this given delivery number. Quick corrections must be made to save time and not have the pro- jected delivery date passed (RightInformation, 2024). 2.8.2 Delivery Management Delivery management highlights one of the most important steps within the supply chain when considering customer satisfaction. It is the process of planning, coordinating as well as tracking the movement of goods through the last part of the supply chain. The main benefits of delivery management and investing in them are the following: faster delivery, lower cost per delivery, increased profitability, as well as increased customer satisfaction (Skyes, 2023). Many companies, like case company 1 as well as 2, tender to send consumers around the country as well as to some neighboring countries. There are several reasons why a transport company used for delivering parcels to the same zip codes should be used around the year. It will save time, reduce expenses, maintain good customer service, as well as guarantee seamless tracking of deliveries. Smaller packages can be set to be picked up by a delivery service provider on set days, which will in the long run save 22 money as bulk deals can be arranged. Using the same delivery companies can ensure safe and secure delivery of goods, as quarterly/monthly meetings are held where key performance indicators are tracked (Farlow, 2024). 2.8.3 Customer Management Customer management or otherwise known as customer relationship management (CRM), represents strategies, practices, as well as technologies to analyze and enhance customer relationships. Efficient customer relationships help build trust between the consumer and company producing the goods, as well as helping improve already known processes to be even better and more functional. A key factor is gaining feedback and looking at what can be already improved to mitigate possible risks or mistakes being made. Many companies do not use a customer relationship management tool, and many the at the same time use one but fail as it does not align with the scale of the company's operations. Using a tool can help analyze and visualize data, support workflows, communicate better with customers, as well as streamline feedback. Having all these tools in one can help both parties perform better. Lastly, using a CRM tool can help traceability, all data is kept in one place (ActiveCampaign, 2024). 2.8.4 Sales Management Sales management can go hand in hand with demand forecasting of production, ware- housing, as well as raw materials in general. Understanding future seasonal trends for example can help provide a clear view of the number of sales that needed to be done. This can be in many cases justified by previous years' seasonal sales as well and com- pared to current consumer consumptions. Sales management plays a key role in under- standing how much production can be produced for a certain product, as that is what they base their sales force off. Sales and operations planning process can be divided eas- ily into a structured approach, starting from workspace preparation to the final plan 23 finalization and release. Within this we have demand planning, supply planning, as well as planning of the review and reconciliation. Identifying bottle necks, cost service levels, and even safety stock can all enhance sales management effectiveness (AnyLogic, 2023). As shown in figure 1, there are many different factors that play a key role in utilizing a ERP software to its fullest potential, which allows sales management processes to flow with ease. 2.9 Comparison between production and warehousing demand forecasting There are many similarities between production and warehousing demand forecasting, both following the same end goal, which is to have enough stock for the next supply chain cycles step to successfully execute. Data collection, forecasting techniques, as well as collaboration within stakeholders all have a crucial role when ensuring each step has been done for all expected outcomes to occur in the supply chain. Demand planning involves forecasting future customer demand and trying to optimize inventory levels and minimize costs around warehousing with overstocking. Many key factors need to be taken into consideration at this point for example variability in demand, disruptions along the way with raw materials before production, or even the complexity of managing different production lines making various products and that all go according to the orig- inal plan (restack, 2024). Production and warehouse demand forecasting can be easily split into two different ar- eas: demand planning and supply planning. Demand planning underlines the need to understand an appropriate product history, trends in the market, as well as campaigns. Sales drastically increase during campaigns and preparing raw material and resources can be challenging. Enough safety stock must be available in case customers bump up the amount they want within a campaign. This matter has a huge effect on supply plan- ning as the correct business operations, transportation, and coordination, as well as management of resources and capacity must be planned. Whether or not the goods are stored directly from the production line in the warehouse, or either sent directly to a 3PL 24 warehouse or the end customer, all have a major effect on planning allocation and the flow of goods (BlueRidge, 2024). Table 2: Production vs Warehousing recap through a table (Restack, 2024) & (Blueridge,2024) As shown in table 2, there are various similarities between production and warehousing forecasting. Safety stock playing a key role in during seasonal sales effects both produc- tion as well as warehousing. Without having the needed amount of packaging material in the inventory of a warehouse, the production cannot run. Supply planning being a crucial role as well has both the need to understand product history, trends, as well as campaigns. Key components being data collection, forecasting, as well as stakeholder collaboration fall under being two different aspects that both production and warehous- ing need. 2.10 Tools to help forecasting within warehousing Forecasting within warehousing can be very difficult at times if production timetables change within a short time. Various tools such as the ABC analysis, as well as batch track- ing can help subject matter experts within warehouses better optimize, the way they store goods, as well as what they have in stock. Real time inventory tracking through an 25 ERP, as well as utilizing historic data can all help forecasting within warehousing. In case sudden volumes pass the amount that a main warehouse can store, third party logistics providers are great to utilize to store goods that cannot fit. It is highly important also to communicate with customers and first see whether they can pick up goods from a dif- ferent location (Macmillan, 2024). Table 3: Optimizing warehousing forecasting and inventory management (Macmillan, 2024) As shown in table 3, there are various aspects that can be wrapped up as being the best tools for optimization. Using the ABC analysis, batch tracking, as well as real time inven- tory tracking through the ERP as well as historical data are crucial for mitigating any mis- takes and enhancing processes. The need to have access to a 3rd party logistics company in case of need of inventory space, as well as communication with stakeholders to see whether or not re routing and picking up goods from a different location is even possible. 2.10.1 Warehouse ABC Analysis The origin of the ABC analysis used in many warehouses around the globe comes from an 80/20 rule, otherwise known as the pareto principle. 80% contributing to outcomes or outputs, and 20% being from causes or inputs. The ABC analysis is an inventory man- agement technique that determines the value of inventory items based on the im- portance and turnaround of a specific SKU. The most important SKUs of a company are classified as A, B being the second most sold, and C being products that do not tend to move as much (ArRacking, 2024). 26 In many cases, companies tend to use even more of these classifications if they have a wider variety of goods that move and do not, and therefore it is not limited to ABC. Be- low is an example of the type of product, importance, as well as percentage of total inventory (Jenkins, 2023). All minor tools used to help inventory control have a huge impact on the capability of a business to reach accurate demand forecasting. Forecasting demand plays a key and crucial role in inventory control by predicting future customer demand based on previous data. Knowing and using past data amongst a company’s SKUs sold, using the ABC analysis can give great insight for storage capacity. Greater stock control, cost reduction, as well as easing employees when picking goods into delivery zones by saving time all help justify why using the ABC analysis is phenomenal to utilize. Saving time when collecting goods and not having to drive across the warehouse to pick up something that is a fast-moving product can easily increase the amount of stock a warehouse can keep when time is at stake. All in all, adjusting the storage system for each SKU type will vastly leverage space when storing (Oleg, 2024). Taking advantage of and easing this process when considering the ABC analysis within a business, using a warehouse management system will boost its overall potential. Config- uring allocation strategies for exact locations within a warehouse can ease traceability and fix possible human errors that may have occurred when products arrive from pro- duction to the warehouse (Mecalux, 2024). Table 4 : ABC Inventory Analysis & Management (Jenkins, 2023). 27 As shown in Table 4, there are different classifactions per SKU or products sold that can be classified in their own classes in order to best utilize space and efficiency when picking deliveries that are about to leave the warehouse. Class A being the most sold products, class B being the second most sold products, and lastly C products that are not sold so much and can be stored further away from department zones of the warehouse. 2.10.2 Batch Tracking and coffee traceability Traceability within the coffee industry is one of the most important factors when looking at the lifecycle from raw materials to finished goods. Arabica as well as robusta or only two examples out of the many different coffee beans that are used amongst different roasteries around the globe. Within production and making the same SKU of coffee from week to another, it is crucial to use the same beans and recipe to maintain customer satisfaction and the same taste. There are 124 different species of coffee, but only two account for 95% of coffee consumed worldwide, robusta as well as arabica (Jordan, 2023). Once the coffee has been classified into a specific product and produced, there is a batch and year that is traced by the warehouse as well as production. Batch formatting in coffee is usually Month as well as year combined, for example 462024. Coffee roasters have a lot of incentives to use coffee that can be easily traced all the way to the country of origin. From a consumer perspective, coffee is something that many are curious to know more about, rather than just drinking it. Roasters are only able to trace and have transparency with coffee that is purchased by buyers that track every- thing from the origin farm that it leaves from (Hermanos, 2022). Without batch tracking, inventory control can be almost impossible within end warehousing processes. Under- standing forecasts and placing sales orders against current inventory in stock is based on tracking finished goods through, for example, an ERP system. Replenishments for the same week's coffee that is made do not occur as they change from one week to another depending on when the production cycle was complete. Understanding unit measures of sales, pallets that the coffee is sold on, as well as which batch a consumer wants only eases the different processes within a coffee company supply chain structure. Sales and customer service have clear visibility as to what goods need to move if they have been 28 sitting, as well as which areas to allocate new batches within a warehouse (Flowtrac, 2022). 2.10.3 Inventory turnover analysis The inventory turnover analysis is a financial metric that often indicates how often a company sells and replaces inventory within a specific time frame. By dividing the total number of days in a given fiscal year by the inventory turnover ratio, a company can determine the average number of days it takes to sell its goods. The inventory turnover ratio provides valuable insights that can assist businesses in making informed decisions regarding pricing, marketing, as well as procurement. This is a key efficiency ratio that companies use to assess how effectively they utilize their assets. Shown below is a cal- culation that can be used to count the inventory turnover formula and calculation to get correct inputs for the analysis (Fernando, 2024). Inventory Turnover Formula and Calculation. Inventory Turnover = COGS / Average Value of Inventory where COGS = Cost of goods sold Understanding the inventory turnover is a great point to show high turnover rate for seasonal goods and usually signals strong customer demand and high satisfaction. This can also rase different issues, with for example, potential problems regarding underpric- ing or stockouts that can be addressed in time before it is too late. Inventory turnover analysis can be broken down into different factors depending on how it affects it and shows the importance of demand forecasting in both production as well as warehousing. It helps monitor efficiency, optimize stock, improves cashflow, tracks trends, and is a great tool to identify possible operational issues (Schneider, 2024). 29 2.10.4 Third party logistics warehouses By utilizing advanced technology such as ERP and warehouse management systems, third party logistic providers can offer real-time inventory visibility which is essential for preventing stockouts, improve demand forecasting, as well as help leverage stock if the main warehouse is reaching its full capacity. Third party logistics parties can manage much more than storing stock, as they can help with freight forwarding, coordinating with carriers, as well as optimizing deliveries. All of these have a huge effect and can help when forecasting peak seasons throughout the year. They can also enhance flows with the help of value-added services which frees space from main warehouses and help lev- erage and ease forecasting (Uribe, 2024). Partnering with a 3PL provider can help businesses significantly reduce the risk of inven- tory errors, delays in transportation, as well as improv overall supply chain visibility, which is crucial. Effectively imaging these aspects enables companies to minimize gaps between supply and demand, as well as enhance the speed and accuracy of order fulfill- ment and ultimately drive business growth (Reid, 2023). 3PL integration can be utilized to its fullest potential by having both the outsourced warehouse send reports of stock levels as well as trigger replenishment orders to bal- ance out stocks. If a primary warehouse is close to full, notifications in the system can trigger a transfer order to an outsourced warehouse to leverage space. Utilizing 3PL part- ners to store goods that move in bulk is a great way to free up space from a main ware- house where there are goods that are not categorized as A products, otherwise known as the most sold products. Fast delivery, scalability and flexibility, as well as end to end warehouse management for accurate stock all are great assets of using them (Hopstack, 2024). Many companies tend to leverage their stocks by sending their most sold products to an outsourced warehouse directly from the production line. In this scenario, full trucks leave the warehouse to the 3PL, from which the retailers can pick up in bulk. This can 30 highly effect balancing resources for the primary warehouse as it can be closed during certain hours, not having both warehouses open and have cost overruns. There are many steps that can slow down setting up a 3PL partner, however, it should never be the case to not use one. Data integration and system compatibility as loss of control and visibility from time due to IT errors can happen. It is highly important to always have integration work as the sales happen from the ERP data, and not from what is physically in stock and in the outsourced warehouses 3PL warehouse management system. Both systems must have the same stock as retail orders are only made from one system, causing problems at times if certain batches are placed on the order but are not physically in stock when it is time to pick up the goods from the outsourced warehouse. This shows the im- portance of cross-checking stock daily and not only relying on the system. Receiving daily stock reports and comparing it to the stock in the ERP can drastically mitigate possible issues that may rise (Symspon, 2024). 2.11 Tools to help forecasting within production Production forecasting is simply the practice of predicting the future demand for retail products and the necessary resources to produce them. These resources typically re- volve around labor, capital, equipment, as well as raw materials. Effective production forecasting enables product-based companies to keep inventory levels balanced, ensur- ing that they can satisfy customer demand while maximizing profitability (Pianaar, 2023). Various methods can be used to enhance and boost the efficiency and accuracy of fore- casting within production. Using trend projection, lifecycle analysis, as well as the mov- ing average method subject matter experts can pinpoint and mitigate to a certain extent main variables that can lead to supply chain disruptions. Within the coffee industry, sea- sonal sales mark a huge benchmark in the annual sales. If experts have not perfected reliable methods for predicting the future, manufacturing forecasts will continue to have areas for refinement. Factors like sudden changes in consumer preferences, unpredicta- ble season fall for coffee, as well as internal issues make it harder than ever at times to predict what should be made, from which batch, as well as at what point in time (Jenkins, 2023). 31 Table 5 : Tools for helping forecasting within production (Jenkins,2023) & (Pianaar, 2023). 2.11.1 Trend projection Within forecasting there is not one correct tool used to predict sales or demand, but rather many ways of proceeding according to what the business must utilize and make judgments off. Trend projection is a forecasting method used to predict future events based on historical data. Depending on the industry, seasonal sales tend to fall under different quarters of the year, however, in coffee it tends to be during certain times of the year. During the summer consumers drink coffee less as it is warmer outside, and then when its temperatures drop, coffee consumption tends to drastically increase. Sea- sonal forecasting falls directly under the trend projection as businesses can easily extract previous data and make assumptions off them. With the data used, subject matter ex- perts forecasting can assess patterns in sales cycle in and know when to take business critical actions. Adjustments must be made on an annual basis as new products reach a market that may not have been in the past. Within production, balancing the main SKUs sold, as well as including projected sales for new products or those that sell less can be difficult at times (Flibier, 2024). Within the time trend projection, removing duplicates, adding missing data, as well as removing low sales figures, it is crucial to not making the wrong forecasts. Within this, short term as well as long term trend projection can be implemented. Short term is fo- cused on predicting how trends, especially micro-trends will evolve soon, typically over the next month, quarter or season. Opposite to that, long term trends look at the poten- tial of progression of macro-trends over extended periods. Within the coffee industry, 32 the main products of businesses sold are looked at from a long-term perspective, whereas the smaller sold products are looked at in a short term (Gaillot, 2023). 2.11.2 Lifecycle analysis method The lifecycle analysis is being implemented amongst different business sectors around the globe more and more each year as they try to become more sustainable. There are certain certifications a business can get, which will help visibility and enhance customer satisfaction. There are various benefits of using the lifecycle analysis as it can help im- prove product development, environmental communication, as well as enhance strate- gic planning for the future of the production cycle and end products made. The market- ing department can get a vast amount of knowledge of factual data that can be commu- nicated onwards. On the flip side, a purchasing department can help learn and adapt to better suppliers that follow sustainable acts. The international organization for standard- ization (ISO) has two different certificates it can provide, 14040 as well as 14044. To fulfill the requirements and to reach expectations, four different factors come into play. Goal and scope definition, inventory analysis, impact analysis, as well as interpretation (Gold- steijn, 2022). There are certain factors revolving around this which can be seen in Figure 2 below. Disposal or recycling, raw material extraction, use, distribution, and lastly man- ufacturing are all ways of analyzing the lifecycle management and are all as highly im- portant as one another. 33 Figure 2 : Life cycle analyses method Diagram (Golsteijn 2024). 2.11.3 The moving average method In statistics, a moving average is a method used to analyze data by calculating averages for different subsets of the entire data set. In production, this method works by taking the average of a certain number of past observations such as stock price over a certain time and calculating the figure into the near or far future. This analysis extracts possible fluctuations and only keeps those that have a stable flow. To get precise analysis, the shorter the time frame revised will give more accurate numbers than those taken from a longer time. More data from a smaller time, or even a longer one will take out possible trends that do not need to be accounted for as they have not occurred many times. There are some factors that need to be added, for example future known events or sales that have already been held. Basing data off previous events and not taking those that are fixed for a certain time that have not occurred, may disrupt the outcome of the moving average method utilized (Akyurt, 2024). 34 Using this method can easily pinpoint certain issues that may have occurred that have not been realized before. This is a great method used when looking at seasonal sales, however, having to consider new materials, production lines, or possible set downtimes. All in all, using this method helps clear out price data, as well as it can easily be adapted on the go. There are a few disadvantages using this as data is taken from the past and does not include current prices. Knowing possible prices changes, for example in the coffee industry and in production, must be taken into consideration to get closer and accurate data. As coffee is a very volatile commodity, this way of analyzing data can be very precise if future approximate figures are known and can be considered (Wong, 2024). 2.11.4 Time series analysis Time series analysis links to many of the other analysis methods already written about within this research. Learning the correct model to enhance and use plays a key role in the result. Learning the correct model, and the algorithms used that we use on sample data can be used several times to make different outputs with minor adjustments. Time series forecasts cannot rely on traditional validation methods, as well as preventing bi- ased results can be mitigated by using observations that precede those in validation tests. Once the correct method and model is chosen, it should be trained on the full training set with its performance evaluated on a separate test that comes after the training data time. This method of forecasting can be used just like any other ones that has been al- ready mentioned, especially for predicting seasonal consumer demand for a given prod- uct. It can utilize data to predict stock prices and can be done using either statistical models or machine learning techniques (Infludata, 2024). 2.12 Enterprise resource planning and manufacturing execution system In current markets and businesses using both an enterprise resource planning and man- ufacturing execution system is extremely important if volumes produced are high. They ensure that all production-related activities align with supply chain management sets, preventing and showing a clear view of current stock, future demand, as well as ware- housing capabilities. Using both systems provides enhanced resource allocation, better 35 decision making, as well as improves overall communication between different depart- ments of a business. Depending on the size of the business, there are many standards and regulations that must be met by the authorities. Compliance with certain standards for auditing reasons can easily be traced using a MES or ERP system to extract data from a certain timeframe (Morrisson, 2024). Quality assurance plays a key role in the coffee industry and therefore certain bench- marks can be tracked using these systems. Integrating both ERP and MES to talk with one another increases visibility and gives subject matter experts time to react if something needs to be changed. Through automation, human errors can be drastically minimized. Automated systems guarantee the timely and precise collection of data from production units and work centers, increasing the reliability of data. This precise method is crucial for building trust in the system and supporting correct assumptions (ProManage, 2024). 2.13 Demand forecasting SWOT analysis To understand the strengths, weaknesses, opportunities, as well as threats within de- mand forecasting it is vital to break these down and open each sector to understand how they influence one another. Without proper demand forecasting a company cannot run the supply chain in a structured matter, which leads to sales not met as well as cost over- runs and no customer satisfaction. Strengths under the use of structured demand forecasting - For example, maximized inventory management, simplified labor hiring and scheduling, improved cash flow, as well as reduced storage over or understock (Blueridge, 2024). On the flipside, there are several weaknesses in demand forecasting that subject mat- ter experts face annually: - Data complexity being one, where finding the correct outcomes from a large data set can lead to bad assumptions. Fortunately, artificial intelligence has improved over the past years but cannot be relied on fully due to it being able to make mistakes as well. Secondly, different markets being very volatile and prone to 36 being affected by market trends and raw material pricing. New product launches, consumer purchasing power, and even competitor pricing can all negatively af- fect the outcome when the product is done (Brown, 2024). There are various opportunities that a business can benefit from when diving deeper into the world of demand forecasting: - Understanding market trends, customer purchasing power, as well as mitigating and minimizing ongoing recurring issues. If a business lacks visibility and does not monitor, implementing better demand forecasting tools can open doors to prob- lems that have not been seen in the past and stop them from happening again. Optimizing supply chain operations and giving more accurate data for future op- portunities of a business are all great perks in forecasting (Kaushik, 2024). Certain threats and challenges always exist when considering demand forecasting - Data inaccuracy and availability can lead to wrong moves within production and warehousing, affecting the finance and overall sales of a company. Whether the threats come from neglecting external factors, overlying complex models, lack of communication, or failure to update and adapt to change, all factors have a dom- ino effect on one another if not executed correctly (Sharma, 2024). 2.14 Importance of silos and monitoring stock for coffee pre-production Using silos as storage for coffee pre-production in the short or long term is vital for a business within the market. Geographical problems for coffee producers in, for example Europe must purchase and inbound coffee weeks before production as there are certain steps that need to be conducted. Checking each batch from the container that arrives, tasting the coffee, as well as classifying it are all steps within the coffee supply chain that must be done with all beans to the plant site. Efficient use of vertical space with silos being placed side by side allows thousands of kilos of coffee to be stored side by side, not taking a large geographical space. Coffee bean silos can usually accommodate sev- eral tons, catering to the needs of large-scale coffee producers or suppliers. Once coffee 37 has been classified by the quality assurance team, easy access to coffee near the pro- duction site can mitigate shortage of raw materials as well as provide resources when there is sudden increase in need for production to sales (Coban, 2024). Having an ERP in use where stock is monitored within the silo, as well as proper integra- tion with the MES system from production, a business enhances traceability by integra- tion between the new systems. As soon as the coffee is sent from the silos to the roasting phase, a message is sent within the system to the MES for the production order. In addi- tion to storing coffee under optimal conditions, roasters must also ensure that roasted coffee beans are handled and transported properly and safely within facilities. Therefore, it is crucial for silos to be close to the production site to mitigate any external factors that may affect the quality of the beans (PerfectDailyGrind, 2023) 3 Methodology and research approach A qualitative approach was used in this thesis rather than a quantitative as the re- searcher wanted to focus on a smaller sample size group (10 interviewees for the inter- views section), to get detailed in-depth responses from those working for the case com- panies presented. This amount was decided since we could get one person from each department of the supply chain and see their intake on how demand forecasting effects their daily operational tasks. The focus was on two case companies, and therefore this could be justified as enough interviewees as it had every subject matter expert in the scope that should be. This approach allows the researcher to flexibly gain information from the ones interviewed and ask why and how the answers were stated the way they were. By using this technique, one can understand and gain possible customer prefer- ences from subject matter experts, gain knowledge on possible market trends, and draw key findings from those that work in the industry daily. All in all, a qualitative approach allows the researcher to gain knowledge on how and why certain actions are done and proceeded with. It allows the researchers to catch onto any changing actions amongst the target group. It is a much more flexible approach than quantitative approach and can target key factors that need to be monitored (Tim Vaughan, 2021). 38 Throughout this research, a methodological approach was focused around three main factors. Exploring expert judgement, understanding organizational processes and prac- tices within forecasting, and investigating the impact of uncertainty and variability all had a crucial role. To properly execute and find key findings, as mentioned above, a struc- tured interview approach was developed. Trending topics, gaps within a department, as well as main key points could be easily extracted by having structured interviews, rather than open-ended ones. Through structured interviews, the main target questions could be directly asked rather than receiving possible misleading answers from not so struc- tured questions. A key benefit of using a qualitative approach within the research is its capability to not only take previous data into consideration, but also enables one to anticipate how spe- cific variables might affect sales cycles. The main reason for this choice was the fact that qualitative analysis on a case two case companies and business area focus on experts’ insights. If data were to be taken from the past it may be outdated. Interviewing subject matter experts in real time and analyzing the data gives the correct updated and ongoing issues that may be occurring. Looking at a span of data before the covid pandemic as well as wars that are going on, and analyzing that, can give very misleading information to a reader in the current market. Understanding ongoing issues and trends is vital to understand the scope and way demand forecasting works currently around the globe. 3.1 Case Study Companies The case study is focused on a coffee company that has operations as well in a smaller factory (case company 2) one hour away. The reasoning behind choosing both case com- pany 1 as well as 2 even though most forecasting is done from case company 1 is the fact that it is good to see how forecasting methods are different when the volumes are dras- tically higher in one or another. The company provides different coffee, both organic coffee as well as regular coffee. 39 Case company 2 provides to smaller suppliers and coffee that they want to design them- selves, whereas case company 1 focuses more on the retail side and consumer coffee that is sold in larger amounts. ABC classification is used in case company 1, where the most sold SKUS are classified as A products, second most sold products as B, as well as C products which are not sold so much. Both companies sell grinded coffee as well as beans that those purchasing can grind themselves. Different sizes of packages can also be purchased, as all are not the same size. Case company one employes hundreds of employees from floor level workers all the way to those working in offices around the country. Case company one also provides coffee machines to companies that are willing to purchase them for different purposes. The companies have been in the industry for tens of years and have vast knowledge on coffee supply as well as production. Certain forecasts can be done without tools if pro- duction levels are much lower, which is the case in the second case company. As soon as volumes are higher as well as the number of various products varying weekly in produc- tion, certain tools must be used to forecast as the amount of data needed to be revised before production starts is extraordinarily big. 3.2 Data Collection The collection of data used within the case study is dummy data (found in the last part of the thesis) due to confidentiality constraints that may occur if real data were to be used. All data is based off something, but the actual outcome and numbers shave been changed. The case study companies used (1 & 2) are based off real companies that work hand in hand within the coffee industry. The broad interviews conducted are towards subject matter experts within the field that work daily within sourcing, logistics, produc- tion, and many other parts of the complicated supply chain. The interviews are kept anonymous, however the title of the employee interviewed will be stated. Further infor- mation in time of interview, years of experience, as well as departments they work for can be found below: - Several years of experience for each interviewee (ranging from 5 years to 40 years in the industry). 40 - Each interview took roughly 20-25 minutes - Departments included: logistics, integrated business planning, production, fi- nance, warehousing, sourcing After interviewing different employees (answers found in appendix) from various teams within the supply chain, certain trends popped up. On time communication came up in almost every answer, where planning horizons varied from a few weeks up to even six months. Within capacity and budgeting, warehouse capability for storage played a key role when considering situations where overstocking demand may need to be done. Looking at demand forecasting from a bird's eye view from a sales perspective all the way to the end of finished goods, each supply chain procedure and process must under- stand what is happening before as well as what is going to happen next. Time is crucial when having a set goal for a consumer to have their products, and therefore immediately informing other departments if changes are made can save unexpected changes from escalating. 3.3 Discussion and recommendations The importance of technology in demand forecasting is evolving each year. Enabling us to downsize the amount of subject matter experts working on everyday forecasting will likely occur soon. Analyzing data with different tools such as Microsoft office excel, ERPs or other programs can provide great insight in a few minutes with fixed parameters. Technology also plays a key role in GPS tracking of container freight, to let workers know lead times or any changes in the transport timing. While using 3PL partners for storage place for goods that are moved more in bulk, it is important for the coffee company to always be on top of stock levels in outsourced warehouses. Certain integrations and daily reports that are sent from the system can help compare stock, however, with the use of warehouse management systems, a worker can go into the system and see livestock and adjust in the ERP if needed. Without proper corrections done on time, it is a domino effect if batch to batch transfers are made in the ERP and not correctly corrected. 41 3.4 SWOT analysis around structured interviews Throughout the interviews conducted, there were certain trends that came up several times. Various strengths, weaknesses, opportunities, as well as threats came up which are good to elaborate and make an analysis from. Great communication between stake- holders came up as one of the key elements spoken about. Daily or weekly meetings between different departments were shown to enhance transparency as well as the ca- pability to seek problems on time prior to them causing negative consequences within the supply chain. Great traceability as well as transparency between MES as well as SAP was addressed by main of the interviewees, giving each department a clear view on how the supply chain works. Quick adaptation to changes was one of the weaknesses brought up, not allowing enough time for changes to be made in the next steps of the supply chain cycle. Stock allocation and storage space issues were fortunately easily resolved with the use of 3PL partners were stock can be transferred. Campaigns being one of the main sale scopes throughout the year, causing issues at times with amounts changing in a short time. With close collaboration with different stakeholders and departments, issues were resolved quickly, not leading to larger problems. Various opportunities were addressed with the implementation of different tools to check statuses of current processes. Using warehouse management services, the ERP system, as well as the MES numerous different problems from previous busy seasons were brought up and mitigated not to happen in the future. Great communication re- volving around demand forecasting and resources needs were seen as an opportunity as well to improve possible flows to make them even more visible to different stakeholders and enhance the raw material to finished goods process. The coffee industry being very volatile with pricing, certain threats came to play when considering demand forecasting. Disruptions with freight forwarding of coffee from the origin country can easily prolong the projected schedule for production to start for the 42 given batch. Quick changes with retail buyers for ready made goods can also affect sales with coffee being produced based on forecasts. Therefore, clear communication be- tween sales department as well as retail buyers is vital to adapt to changes on time, prior to it back too late. 3.5 Analysis of sample data The following chart below can be seen in the last part of the research as an extension as well. The example above is a hypothetical and created example of a 5-week span with differ- ent scenarios of forecasting in coffee. It is critical to ensure the right quantity of beans is roasted and ready for distribution. The overall example above is presented from a exam- ple where retail buyers make adjustments too late, causing disruptions in the supply chain. In week 1 we can see that the forecasted demand was 5000kg, however, the de- mand was increased by five hundred kilos. Late notice buys retailers purchase decisions did not give enough time for production to react, causing a shortfall of 500kg produced. This example shows that almost all sales were met, however adjustments made too late did not leave opportunity to adapt to production amounts. 43 During the second week there was slightly a different point of view taken as 5200 was the forecasted demand, but the actual demand was slightly lower at 5 tons. The retailer did not make changes now, leading to an additional amount of 200kg left to be stored. Fortunately, coffee is not something that goes bad quickly, leaving margin for the given weeks batch to be sold later. Different storage issues due come to play as this was just the production forecasts. Warehouse forecasts did not have this in consideration, caus- ing possible issues with storing as it was not taken into consideration. Fortunately, the additional volume produced that was not sold was not so high, leaving space for the warehouse to allocate it on time without a problem. Week three demonstrated the more severe impact of late adjustments. Forecasted de- mand was 5,5 tons but retailers increased their order by 700kg, increase the actual de- mand to 6 tons. Adjustments were made in production; however, the response was too late. As a result, the production team had made a half ton increase which then needed to be stored and used elsewhere. Additional costs overruns came to play, leading to a bad view on the supply structure of this week. Week 4 forecasted demand was 5,8 tons but the actual demand drastically decreased to 4.2 tons, creating a large shortage. Re- tailers adapting to change this late caused a huge disruption in the roaster’s operations. This all had a huge impact on unfilled orders, delayed shipments, as well as sales losses. Poor communication here was the root cause that had the major effect on the outcome. Week five showed us a flipside to the other scenarios, where the forecasted demand was 5 tons, when the actual demand increased by three hundred kilos. The retailer buyers had adjusted their order upwards, which was not a huge increase from the previous week. The roastery adapted to the change and could change the amount of coffee that was produced. Managing inventory was the main issue here, with all the previous steps being executed on time due to proper communication in a short time frame. 3.6 Sample size calculation and analysis The reasoning behind the sample size only being ten was the fact that the additional unnecessary information from non-subject matter experts did not want to be included 44 in the research. Those working hands on around demand forecasting within production as well as warehousing are the best experts to answer questions that were asked within the interviews given. In addition, none of the employees interviewed were new to the industry or recently joined the business, but rather were working there for many years and have vast knowledge during different peak seasons throughout the previous years. Knowing previous examples before, during, as well as during the pandemic that started in 2019 gave a very clear view on different hectic times that drastically effected the sup- ply chain in many ways. Looking at a specific case company, ten subject matter experts may seem small to some readers, however, a subject matter expert from each depart- ment involved around possible demand forecasting effects was interviewed. To properly calculate the sample size, a standard formula can be used as shown below: 𝑛 = 𝑧2 ∗ 𝑝 ∗ (1 − 𝑝) 𝐸2 N being the required sample size, Z being the Z- Score (depending on confidence level), P being the estimated proportion as well as E margin of error. Using this calculation and basing it off ten interviewees in total, in order to get a 95% confidence level with a 5% margin the sample size should be much larger. This is taking into consideration if the end results were looking for a more generalized view, rather than those given from subject matter experts. Instead, this research seeks to look at qualitative insights from experts and therefore the sample size can be much smaller as it covers the majority of scope within demand forecasting for production as well as warehousing. 3.7 Managerial implications All in all, all the different scenarios mentioned throughout the research have a huge im- pact in one way or another in the supply chain. Whether the consumers get the coffee on time with the minor adaption within the supply chain, certain cost overruns may oc- cur that are not budgeted. Some of amounts of coffee that are produced may need to be thrown away eventually if produced too much and no sales are met. In the scenario from one week where there was too much produced cause a bad view towards custom- ers as they did not receive the coffee that was ordered. Certain situations are very 45 difficult to overcome within the supply chain if there are for example strikes ongoing and freight forwarding of green coffee from ports to the site is on hold. Therefore, always having a different option within the supply chain and creating different risk analysis around possible threats is a great way to mitigate problems in the near or far future. There are various ways of improving on demand forecasting approaches. Improving the quality and granularity of data collected being one main key player. Ensuring the data that has been used is comprehensive and up to date. Comparing to previous year's sales is not always the correct approach if there had been an economic crisis then and not now. Taking sales as well as economic indicators, weather patterns for coffee fall, as well as competitor activity, can all have a major impact on improving forecasting in the near or far future. (Logility, 2023). Correlating historical demand fluctuations with external events may be challenging at times, however, can give much more precise information when forecasting for the future. By combining possible historical data within the company and adding possible historical event data, certain trends may rise that can be considered when forecasting the next cycle. Taking technology into play, utilizing machine learning techniques such as time series analysis, deep learning models, or even predictive data analytics can enhance a forecasters approach on what they are trying to seek and making conclusions from (Pre- dictHq, 2024). Throughout the use of machine learning, there are certain areas of de- mand forecasting that have been uncovered due to its accuracy and capability to crunch data that is so large and would take tens of manual hours of a human to do. Time is crucial within the supply chain, and therefore even if humans could do it, it could be too late due to adaptions and needs of starting production prior to set schedules. One of the main key notes of machine learning is the capability to process and analyze data from various sources at the same time. Whether it is social media, weather forecasts, or even economic indicators, machine learning can give great outcomes within a matter of minutes, something that humans would take days to do (Laskova, 2024). 46 4 Study results and analysis Campaigns play a key role in seasonal sales as well as when looking at total volumes annually. Receiving accurate forecasts from retail buyers plays a key role in easing the work for almost all roles in the supply chain of the coffee company. Without receiving campaign amounts on time, time constraints come into play as there first must be enough green coffee in stock to start production. Certain situations have happened where green coffee transportation was delayed, causing production scheduling to be re- looked at. Fortunately, green coffee from a closer destination was purchased, which eased stock levels and let production flow in a normal matter. Seasonal changes, market trends, as well as economic shifts have an impact on demand forecasting highly. Economic shifts have a huge impact on consumer buying power and whether they will even consider buying anything. Most annual sales come from seasonal trend campaign sales, and therefore having the correct amount of coffee in stock is cru- cial. Since the transit time for green coffee from, for example, Africa is so long, trackabil- ity is essential when waiting for coffee to arrive. Changing routing/ports or even shippers may need to be done if a subject matter expert can see that there is a possible strike, for example, in the upcoming transit harbor. Inflation occurring and green coffee pricing go- ing up, which influences sales pricing can drastically affect buying power due to prices rising. Coffee pricing being very volatile, however, not adjusted very so often (2-4 months), can help forecasters anticipate possible market trend differences. Accuracy within demand forecasting is something every planner strives for. In many cases it is not easy to conduct, as one may rely on SAP forecasts and not receive proper communication from other stakeholders involved. Being proactive and seeking improve- ments outside of the usual flows can help mitigate risks. Therefore, all changes and pro- cess flow changes must be communicated to eat stream down the supply chain to be prepared for them. Larger production can in many cases be dealt with within warehouse capacity, however, resources of workers must be planned of time. There are various 47 variables that may fall in to play and be fine with change, however, many forget im- portant ones that can then be the game changer when it is time to execute. Forecasting has a huge effect on inventory management, stock replenishment, and stor- age decisions within a company's warehouse. Correct product demand must be known in time to arrange transport to our 3PL partners or rearrange layouts in our own ware- house. Some orders may leave late which causes backlogged picking and overlapping of another order that should be picked and placed at a certain loading station. In case of tools not used, warehouse supervisors and workers must quickly rely on changes if re- plenishments happen in hours. Therefore, demand forecasting is affected drastically by inventory management and other sectors, for example fill rates, as production and amounts are based off forecasts beforehand. 4.1 Study findings and recommendations for future managers If one were to hypothetically think of different scenarios that is not within the field of demand forecasting within supply chain, it would be very hard to pinpoint specific ideas that were found in the interviews conducted. Throughout the interviews the reader can gain vast knowledge from subject matter experts on what key factors influence demand forecasting and how can it be more efficient using different tools. In addition, how accu- racy can be looked at rather than receiving any kind of data and making false assump- tions from forecasts. The study can benefit those seeking for better ways of adapting to supply chain disrup- tions, enhance their knowledge on the key factors that play a key role in executing fore- casts properly. Knowing how important informing other departments about changes as well as upcoming possible events that may affect the supply chain cycle is crucial, as it may affect the outcome of the end products production time. Adapting to changes for example with packaging material being late and not being able to produce specific coffee has a major impact on sales, as certain products are classified as A products and sell more. Replacing other products in production can be hard as well due to the simple fact that different coffee blends have different ingredients and bill of materials within them. 48 Fluctuating prices can drastically effect production cycles as coffee prices tend to go up ever so often. Retail buyers get this knowledge and tend to buffer their inventory with the lower prices before the price change happens. This causes a domino effect amongst those buying and a bottle neck for production as many want to purchase coffee at a lower price. Communication yet again plays a key role in this case as it affects each de- partment on role. Logistics must be prepared to change and increase routing, production must have enough resources to produce the coffee, as well as the warehouse must utilize in many cases third party logistics providers to have enough space to store goods. Future recommendations for managers can be easily wrapped from findings found within the conducted interviews. Familiarizing each employee with other departments tasks and how they influence one another is crucial. Understanding if one process is not executed properly and how to communicate it to others as well as mitigate possible sce- narios in the near or far future. One of the most important notes is to let each employee be aware about the importance of informing anything they realize that is not going as agreed as it may not only effect their own apartment but every other one as well. 5 Conclusion This thesis is set out to explore the importance of demand forecasting within production and warehousing operations, with a specific focus on the coffee industry. The primary objective of this research is to determine the importance of demand fore- casting, as well as whether there are certain variables that have a significant impact on decisions being made. How in fact demand forecasting can improve production, as well as whether it can mitigate complications within the suppl chain, as well as how it affects business and customer satisfaction. Throughout the thesis answers to what role does technology play in enhancing demand forecasting processes in production and ware- housing, key factors influencing forecasting accuracy, as well as how does collaboration between stakeholders between accuracy all were answered. 49 The case company(s) used to justify knowledge base within the research was a great choice as it showed two different scenarios where one of the two companies had higher production levels and needed more planning within forecasting than the other. The im- portance of having clear roles within each department working close with one another was a commonly spoken subject in the structured interviews conducted. Certain crucial factors such as communication, understanding market trends, economic shifts, understanding consumer campaign sales, as well as the collaboration and im- portance between stakeholders inside the supply chain all came up as the main and most key areas. Capability to adapt to change and have production work close with warehous- ing to fulfill order management, stock allocation, as well as customer demand all had crucial importance. Using different tools such as ERP as well as warehouse management systems rose many times during the interviews conducted. Throughout the research saw how technology plays a vital role in enhancing demand forecasting processes in the cof- fee production and warehousing sectors. Key factors that influenced forecasts could not be predicted at times since customers tend to change amounts last minute. Collabora- tion with stakeholders as well as clients weekly as well as monthly can always help miti- gate possible supply chain issues as well as improve accuracy of forecasts conducted. Overall, the research revealed the key findings with technology, communication, and un- derstanding market trends and decisions changing within a short time. Also, the study highlighted the significant role of understanding seasonal trends, in this case campaigns, as they play most annual sales. 5.1 Key findings and limitations Many different key findings as well as a few limitations were found throughout this re- search study. Through structured interviews with key stakeholders in the company, in- cluding managers, integrated business planning, finance, warehousing as well as produc- tion, this study gained valuable insights into the company's current way of working. The research found that the company mainly relies on historical data for future forecasts and following seasonal patterns to predict demand. Through the sample data, one could see 50 that the root cause for sales for one product do not always depend on the taste of the coffee. Sudden changes in sales unit type of sales can affect forecasting of overall sales if customers are not willing to buy the coffee in different pallet types. A huge impact on cross-departmental communication came up a numerous number of times throughout the structured interviews. Accuracy being a key term throughout the research, the in- terviews revealed that better alignment between sales as well as all supply chain teams significantly improves the accuracy and precision of forecasting. Specific case company structured interviews, size of interviews, as well as bias and focus on current practices are all limitations found within the study. The study conducted structured interviews with a limited number of key stakeholders, which may have re- sulted in a narrow perspective of the issue. Participants in the interview gave their own view on the answers asked, and therefore this could have led to responses that reflect idealized practices rather than general ones. The case company's current way of working was addressed; however, it did not come across potential future ways of working to mit- igate problems that are current occurring. Lastly, the research was conducted over a lim- ited timeframe, with most interviewees giving similar examples of problems that have occurred in the past months. Having done the interviews over a larger time, other pos- sible answers could have been raised. 5.2 Future research suggestions Further research could focus on comparing the effectiveness of different demand fore- casting models in market with high volatility versus those with smaller ones. This could provide insights into which models are more adaptable or accurate. Since the current forecasts are mostly based on the previous year's sales and models, this could give a totally different intake. Another very interesting topic could be around sustainability con- siderations in demand forecasting for warehousing and production. With the increasing pressure to adopt sustainable practices, future research could show how looking at de- mand as well as trying to be sustainable can have an impact on decision making. Balanc- ing the main factors can have drastically different ways of working when trying to, for example, reduce waste, energy consumption, as well as carbon emissions. 51 6 References ShipBob. (n.d.). Demand forecasting: What it is and why it matters for your business. ShipBob. https://www.shipbob.com/blog/demand-forecasting/ Keepler. (2022, July 4). Importance and benefits of forecasting customer demand. Keep- ler.https://keepler.io/2022/07/04/importance-and-benefits-of-forecasting-customer- demand/ Deskera. (n.d.). What is demand forecasting and how does it work? Deskera. https://www.deskera.com/blog/demand-forecasting/ Flow.space. (n.d.). What is demand forecasting? A guide to the importance and tech- niques. Flow.space. https://flow.space/blog/demand-forecasting/ Maersk. (2023, September 27). Supply chain forecasting: How to forecast demand and improve supply chain management. Maersk. https://www.maersk.com/logistics-ex- plained/supply-chain-management/2023/09/27/supply-chain-forecasting Pacemaker. (n.d.). Demand forecasting meets artificial intelligence: How AI improves forecasting accuracy. Pacemaker. https://www.pacemaker.ai/en/blog/demand-forecast- ing-meets-artificial-intelligence#:~:text=%E2%80%8D-,Accurate%20and%20dy- namic%20forecasts,decisions%20in%20the%20supply%20chain. APUS. (n.d.). Artificial intelligence in retail and improving efficiency. APUS. https://www.apu.apus.edu/area-of-study/business-and-management/resources/artifi- cial-intelligence-in-retail-and-improving-efficiency/#:~:text=Businesses%20can%20opti- mize%20their%20inventory,than%20traditional%20methods%20of%20forecasting. Akkio. (n.d.). Demand forecasting using machine learning. Akkio. https://www.ak- kio.com/post/demand-forecasting-using-machine-learning 52 Dan, S. (2021, November 19). Advancing coffee demand forecasting with machine learn- ing: A data science perspective. Medium. https://medium.com/@sebnem_dan/advanc- ing-coffee-demand-forecasting-with-machine-learning-a-data-science-perspective- 3da09492d416 TransImpact. (n.d.). Forecasting method: Moving average. TransImpact. https://transim- pact.com/nextsights/forecasting-method-moving-average/ Encora. (n.d.). Exponential smoothing methods for time series forecasting. Encora. https://www.encora.com/insights/exponential-smoothing-methods-for-time-series- forecasting#:~:text=Exponential%20Smoothing%20Methods%20are%20a,these%20val- ues%20get%20exponentially%20smaller. Investopedia. (2023, August 28). Autoregressive integrated moving average (ARIMA). In- vestopedia. https://www.investopedia.com/terms/a/autoregressive-integrated-mov- ing-average-arima.asp EazyStock. (n.d.). Inventory forecasting models: Quantitative and qualitative methods. EazyStock.https://www.eazystock.com/blog/inventory-forecasting-models-quantita- tive-qualitative-methods/#:~:text=Qualitative%20forecasting%20mod- els%20are%20based,of%20how%20the%20market%20works. Indeed. (n.d.). Delphi method: Definition and steps for using it. Indeed. https://www.in- deed.com/career-advice/career-development/delphi-method The Corporate Governance Institute. (n.d.). The Delphi technique. The Corporate Gov- ernanceInstitute. https://www.thecorporategovernanceinstitute.com/insights/lexi- con/the-delphi-technique/ 53 Fresh Cup. (2023, February 9). Waste no beans: A demand planning guide for roasters. Fresh Cup. https://freshcup.com/waste-no-beans-a-demand-planning-guide-for-roast- ers/ LeewayHertz. (n.d.). AI in demand forecasting: How artificial intelligence improves accu- racy. LeewayHertz. https://www.leewayhertz.com/ai-in-demand-forecasting/ Dan, S. (2021, November 19). Advancing coffee demand forecasting with machine learn- ing:Adatascienceperspective.Medium. https://medium.com/@sebnem_dan/advancing- coffee-demand-forecasting-with-machine-learning-a-data-science-perspective- 3da09492d416#:~:text=Several%20stakeholders%20are%20using%20machine,predic- tion%20of%20coffee%20consumption%20trends. Leafio. (n.d.). ERP inventory management: How ERP can improve inventory accuracy and control. Leafio. https://www.leafio.ai/blog/erp-inventory Throughput. (n.d.). Inventory management: Best practices and key strategies for effi- ciency. Throughput. https://throughput.world/blog/inventory-management/ Throughput. (n.d.). Inventory management: Best practices and key strategies for effi- ciency. Throughput. https://throughput.world/blog/inventory-management/ GourmetPro. (n.d.). Coffee market trends: Expert insights. GourmetPro. https://www.gourmetpro.co/blog/coffee-market-trends-expert-insights PlanetTogether. (n.d.). Demand forecasting models for production planning in medical manufacturing. PlanetTogether. https://www.planettogether.com/blog/demand-fore- casting-models-for-production-planning-in-medical-manufacturing 54 EazyStock. (n.d.). 8 best inventory demand forecasting techniques. EazyStock. https://www.eazystock.com/blog/8-best-inventory-demand-forecasting-tech- niques/#:~:text=get%20a%20definition%3A-,What%20is%20inventory%20de- mand%20forecasting%3F,stocking%2C%20for%20optimum%20inventory%20control. RightInformation. (n.d.). Order management and demand forecasting system. RightIn- formation. https://rightinformation.com/case-studies/order-management-and-de- mand-forecasting-system/ Routific. (n.d.).What is delivery management?. Routific. https://www.routi- fic.com/blog/what-is-delivery-management Farlow Transport. (n.d.). 5 reasons why you need a transport company for your retail business. Farlow Transport. https://farlowtransport.com.au/5-reasons-why-you-need-a- transport-company-for-your-retail-business/ ClickUp. (n.d.). Customer management strategies: Tips to manage your customers effec- tively. ClickUp. https://clickup.com/blog/customer-management-strategies/ ActiveCampaign. (n.d.). Customer management: Definition and strategies for success. ActiveCampaign.https://www.activecampaign.com/glossary/customer-manage- ment#:~:text=It%20involves%20managing%20every%20touch- point,build%20long%2Dterm%20customer%20loyalty AnyLogic. (2021, May 10). Sales and operations planning in supply chain. Medium. https://medium.com/@anylogic/sales-and-operations-planning-in-supply-chain- 4b2e1c2c1f3a Restack. (n.d.). Model optimization: Answering demand planning vs production planning with AI. Restack. https://www.restack.io/p/model-optimization-answer-demand-plan- ning-vs-production-planning-cat-ai 55 Logility. (2023, April 18). 4 steps to skyrocket your demand forecasting accuracy. Logility. https://www.logility.com/blog/4-steps-to-skyrocket-your-demand-forecasting-accu- racy/ Enago. (n.d.). Research aims and objectives. Enago Academy. Retrieved November 28, 2024,fromhttps://www.enago.com/academy/research-aims-and-objec- tives/#:~:text=Provides%20direction%3A%20Clear%20aims%20and,study%20re- mains%20relevant%20and%20meaningful. Last Mile Logistics Solutions. (n.d.). Achieve efficiency with partnerships. Last Mile Logis- tics Solutions. Retrieved November 28, 2024, from https://www.lastmilelogisticssolu- tions.com/achieve-efficiency-with-partnerships/ Macmillan, S. C. G. (2024, November 28). Essential inventory forecasting tools and tech- niques for e-commerce. Macmillan Supply Chain Group. Retrieved from https://www.macmillanscg.com/blog/essential-inventory-forecasting-tools-and-tech- niques-for-e-commerce/ AR Racking. (n.d.). ABC inventory method in the warehouse: Origin, characteristics, and advantages. AR Racking. Retrieved November 28, 2024, from https://www.ar-rack- ing.com/en/blog/abc-inventory-method-in-the-warehouse-origin-characteristics-and- advantages/ Cleverence. (n.d.). Greater inventory control begins with demand forecasting. Cleverence. Retrieved November 28, 2024, from https://www.cleverence.com/articles/business- blogs/greater-inventory-control-begins-with-demand-forecasting/ Mecalux. (n.d.). ABC analysis: Inventory classification. Mecalux. Retrieved November 28, 2024, from https://www.mecalux.com/blog/abc-analysis-inventory-classification 56 Jordan. (2023, November 9). Where do coffee beans come from? Coffee Spirit. Maxi- Coffee. Retrieved December 2, 2024, from https://www.coffee-spirit.maxicoffee.uk/ar- ticle/where-do-coffee-beans-come-from/ Hermanos Coffee Roasters. (2022, March 29). Coffee traceability: What is it and why does it matter? Hermanos Coffee Roasters. Retrieved December 2, 2024, from https://hermanoscoffeeroasters.com/blogs/blog/coffee-traceability-what-is-it-and- why-does-it-matter?srsltid=AfmBOor79YQsvQHJ-8ctRLqAO6H- If46jFyl0Xo4WLxjr61NmAxH3ssG Flowtrac. (2022, November 23). Coffee tracking: What is it and why does it matter? Flowtrac. Retrieved December 2, 2024, from https://www.flowtrac.com/2022/11/23/coffee-tracking/ Fernando, J. (2024, September 16). Inventory turnover ratio. Investopedia. Retrieved De- cember2,2024,from https://www.investopedia.com/terms/i/inventoryturnover.asp Warehousing and Fulfillment. (2024, May). How to use inventory turnover analysis for stock control. Warehousing and Fulfillment. Retrieved December 2, 2024, from https://www.warehousingandfulfillment.com/warehousing-and-fulfillment-re- sources/how-to-use-inventory-turnover-analysis-for-stock-control/ Uribe, C. (2024, November 13). The role of 3PL and inventory management in modern supply chains. Carlos Uribe Blog. Retrieved December 2, 2024, from https://blogcarlosu- ribe.blog/2024/11/13/the-role-of-3pl-and-inventory-management-in-modern-supply- chains/ Reid, J. (2023). The role of 3PL and inventory management in modern supply chains. DCL. Retrieved December 2, 2024, from https://dclcorp.com/blog/3pl/3pl-supply-demand/ 57 Sympson, E. (2024, August 29). Enhancing inventory management through 3PL integra- tion. Nautical Direct. Retrieved December 2, 2024, from https://nautical-direct.com/en- hancing-inventory-management-through-3pl-integration/ Pianaar, M. (2023). Production forecasting. ShipBob. Retrieved December 2, 2024, from https://www.shipbob.com/blog/production-forecasting/ Jenkins, D. (2023, December 6). Manufacturing forecasting. NetSuite. Retrieved Decem- ber2,2024, from https://www.netsuite.com/portal/resource/articles/accounting/manu- facturing-forecasting.shtml Flieber, L. (2024). Seasonality forecasting: Know exactly when to push sales and when not to. Flieber. Retrieved December 2, 2024, from https://www.flieber.com/blog/sea- sonality-forecasting-know-exactly-when-to-push-sales-and-when-not-to Gaillot, A.-D. (2023, December 12). Trend forecasting and prediction. Meltwater. Re- trieved December 2, 2024, from https://www.meltwater.com/en/blog/trend-forecast- ing-prediction Golsteijn, L. (2022, July 17). Life cycle assessment (LCA) basics. Pre-Sustainability. Re- trieved December 2, 2024, from https://pre-sustainability.com/articles/life-cycle-assess- ment-lca-basics/ ShipMate Fulfillment. (n.d.). Forecasting sales using the moving average method. Ship- Mate Fulfillment. Retrieved December 2, 2024, from https://www.shipmatefulfill- ment.com/post/forecasting-sales-using-moving-average-method#:~:text=The%20mov- ing%20average%20method%20is%20a%20forecasting%20tech- nique%20used%20to,that%20figure%20into%20the%20future 58 Wong, K. (2024, October 8). Identify trends with moving averages. OANDA. Retrieved December 2, 2024, from https://www.oanda.com/us-en/trade-tap-blog/trading- knowledge/identify-trends-with-moving-averages/#:~:text=Advantages%20of%20mov- ing%20averages&text=Trend%20identifica- tion%3A%20They%20help%20smooth,day%20trading%2C%20and%20swing%20trad- ing InfluxData. (n.d.). Time series forecasting methods. InfluxData. Retrieved December 3, 2024, from https://www.influxdata.com/time-series-forecasting-meth- ods/#:~:text=Time%20series%20forecasting%20is%20a,Astronomy Morrisson, A. (2024, November 20). MES vs ERP: Understanding the differences and ben- efits. Top 10 ERP. Retrieved November 20, 2024, from https://www.top10erp.org/blog/mes-vs-erp#:~:text=MES%20and%20ERP%20integra- tion%20ensures,management%2C%20and%20cash%20flow%20planning ProManage Cloud. (2024, April 9). MES and ERP integration. ProManage Cloud. Re- trieved April 9, 2024, from https://promanagecloud.com/mes-and-erp-integration/ Blue Ridge. (n.d.). Benefits of demand forecasting. Blue Ridge. Retrieved December 3, 2024, from https://blueridgeglobal.com/demand-forecasting/benefits-of-demand-fore- casting/#:~:text=Demand%20forecasting%20can%20tell%20you,anticipating%20com- ing%20spikes%20in%20demand. Brown, C. (2024, November 12). Pros and cons of implementing demand forecasting strategies leveraging. PredictHQ. Retrieved December 3, 2024, from https://www.pre- dicthq.com/blog/pros-and-cons-of-implementing-demand-forecasting-strategies-lever- aging 59 Kaushik. (2024). Demand forecasting: Why it's important and how to do it. MeetRecord. Retrieved December 3, 2024, from https://www.meetrecord.com/blog/demand-fore- casting Sharma, G. (2024, March 19). Optimizing demand forecasting: Challenges and best prac- tices. Inside Supply Management. Institute for Supply Management. Retrieved Decem- ber 3, 2024, from https://www.ismworld.org/supply-management-news-and-re- ports/news-publications/inside-supply-management-magazine/blog/2024/2024- 03/optimizing-demand-forecasting-challenges-and-best-practices/#:~:text=Chal- lenges%20in%20Demand%20Forecasting,can%20impede%20the%20forecast- ing%20process. Coban Silo. (n.d.). Main advantages of coffee bean storage silo. Coban Silo. Retrieved December 3, 2024, from https://www.cobansilo.com/main-advantages-of-coffee-bean- storage-silo.html Perfect Daily Grind. (2023, June). Storing and transporting roasted coffee. Perfect Daily Grind.RetrievedDecember3,2024,from https://perfectdailygrind.com/2023/06/storing- transporting-roasted-coffee/ Joiner, S. (2023, September 21). What is supply chain forecasting? o9 Solutions. https://o9solutions.com/articles/what-is-supply-chain-forecasting/ Vaughan, T. (2021, August 6). 10 advantages and disadvantages of qualitative research. Poppulo. https://www.poppulo.com/blog/10-advantages-and-disadvantages-of-quali- tative-research Owczarek, D. (2023, April 22). Supply chain forecasting challenges. Nexocode. https://nexocode.com/blog/posts/supply-chain-forecasting-challenges/ 60 Indeed Editorial Team. (n.d.). Quantitative vs. qualitative forecasting: Pros and cons. In- deed.https://www.indeed.com/career-advice/career-development/quantitative-vs- qualitative-forecasting-pros-and-cons Noble, J. (2024, May 24). ARIMA model: The basics of time series forecasting. IBM. https://www.ibm.com/think/topics/arima-model Field, D. (2023). ERP inventory system: How an ERP system can improve inventory man- agement. Leafio. https://www.leafio.ai/blog/erp-inventory-system/ Mecalux. (2024). Demand forecasting: How to forecast demand for your warehouse. Mecalux. https://www.mecalux.com/blog/demand-forecasting Jenkins, A. (2023). ABC inventory analysis. NetSuite. https://www.netsuite.com/por- tal/resource/articles/inventory-management/abc-inventory-analysis.shtml Hopstack. (2024). 3PL advantages and disadvantages. Hopstack. https://www.hopstack.io/blog/3pl-advantages-and-disadvantages Golsteijn, L. (2024, March 7). Life cycle assessment (LCA) basics. Pre-Sustainability. https://pre-sustainability.com/articles/life-cycle-assessment-lca-basics/ Jenkins, A. (2022). Order management. NetSuite. https://www.netsuite.com/portal/re- source/articles/erp/order-management.shtml#:~:text=Effective%20order%20manage- ment%20is%20important,inventory%20inaccuracies%2C%20and%20cost%20manage- ment. Laskova, A. (2024). How machine learning is redefining demand forecasting: 6 key meth- ods. Futuramo. Retrieved from https://futuramo.com/blog/how-machine-learning-is- redefining-demand-forecasting-6-key-methods/ 61 7 Appendices 7.1 Research Questions 1) Can you describe how demand forecasting is integrated into your daily opera- tions? Demand Planner: - Everyday tasks involve demand forecasting within this position - At time we look at closer demands, and at times further ones - Closer forecasts are looked at more precisely - Current month forecasts are mostly kept, however, at times they may fluctuate if customers make larger orders in a short time scope - Mostly, forecasts are looked 3 months ahead or even further. Warehouse Manager: - Capacity planning within our own warehouse as well as 3PL partners. Coffee is stored in large amounts in 3PL partners. Correct communication on time to let them know what is coming in the near or far future. - Budgeting also plays a key role in this position which all relates to capacity. The key role is to not store goods for a long time but rather have them move faster. - Understanding and receiving the correct amounts and timing for campaigns (in- formation that comes from the Master supply planner) and using that infor- mation to inform our outsourced warehouse partner about amounts. Master Supply Planner: - This is integrated into everyday work. All forecasts must be on hand in well looked at, otherwise production will not be done on time - Raw material in general must be planned on time, not only for machine use but also for employee resources. - Correct forecasts now have a big impact on future products that are going to be in production in the near or far future. All have a big impact on sales and how we will bring new products to the market. Warehouse Specialist: 62 - No direct demand forecasting in this position, however, correct communication on capacity differences what may suddenly be needed in the warehouse. For ex- ample, information on production amounts to understand do we need to move some goods to our outsourced warehouse because of capacity limits within our main warehouse. - This all has an impact on resources as well as stock - Weekly meeting with production and other stakeholders to understand what fu- ture amounts are going to be made to prepare for what is coming up. - Communication - Understanding sudden changes in the near or far future. Transport and Customs Manager: - Everything evolves around demand forecasting to anticipate what is going to hap- pen in the near or far future. - Volume and amounts are always looked at within transportation to make deci- sions. If we do not have that current information, we compare it to previous data that we have carried. Supply Chain Operations Manager: - Looking at demand at a monthly leveling, which may change. Manual changes may occur during this time as well. - Monthly process- weekly granularity. Supply Planner Packaging Material: - Packaging material supply planner looks at a 6-month scope - Purchase orders are placed 2-3 months prior to shipments being made Business Controller: - Always look at a sales perspective, after which we can dig into the amounts needed to be bought by purchasing. - Close collaboration with key account managers and communication plays a key role. Transport and Customs Coordinator: 63 - Linked with different areas and close collaboration with sourcing to know which containers are coming at what given point in time. - If forecasts are not met for coffee leaving the origin countries, planning transport is very hard. Lead times are long for container transport and therefore this is highly important. 2) How do you perceive the role of demand forecasting in the overall supply chain process, specifically within the coffee industry? Demand Planner: - Using tools to check demand are used to perceive demand forecasting within the coffee industry - Communication with different teams is a key role in this position, to let everyone know about resources needed in the near or far future. Within the coffee sector, campaigns play a key role in forecasting. These are the key drivers in sales. - This is not only a downstream overview (green coffee, packaging materials etc) but also an upstream one. Finance and sales need to see that their set goals are met. Warehouse Manager: - Once a month involved supply review meeting where future campaigns and sales are spoken about. This gives a good insight into what information needs to be passed on to our external warehouses as well as our outsourced internal ware- house provider. Understanding short term demand, as well as even 6 months ahead. Communicating with integrated business planning weekly as well helps perceive the role of demand forecasting in this position. Master Supply Planner: - Correct communication with other teams within the company to correctly fore- cast and have production time. - R&D + all other aspects to correctly have the demand in the future. Warehouse Specialist: - Checking stock and realizing that some product is out of stock in our system within a 3PL warehouse, what they physically do not have. Correct 64 communication with integrated business planning to see when a certain supplier will be able to replenish stock. - Through stock monitoring perceiving and realizing gaps within supply chain (warehousing and stock). Transport and Customs Manager: - Overlooking amounts in the future are looked at from one harbor to another when looking at green coffee transportation. These are looked at one year ahead. - Demand and supply do not always meet when considering transportation. Allo- cating specific amounts for the future for shipment is settled and discussed with suppliers, however, these may change as well. If demand is wrong, and more is needed, finding resources for transportation can be hard at times. Supply Chain Operations Manager: - Understanding the market and fluctuating to changes. Supply chain processes may be hard to change adapt to with a short notice. - Order management is only a few days ahead. - The supply chain lead time is 6 months, delivery comes within a few days. We depend on forecasting accuracy as well a green coffee + stock capacity. Supply Planner Packaging Material: - Projected amounts needed come from Sales as well as then amounts that are in needed in production - Everything is used within the ERP. Business Controller: - Overseeing the birds eye view of projected sales and budgeting. Do we have enough demand to reach the goal that was set. Close collaboration to see that the full flow works from sales, to demand planning, production, as well as ware- housing. Transport and Customs Coordinator: - Close collaboration with different stakeholders, as well as keeping up to date with changes for ocean freight. 65 3) How does demand forecasting influence your specific area (production, logis- tics, planning, transportation) as well as decision-making processes? Demand Planner: - Demand planners’ goal is to ease the tasks that are happening afterwards. To know what is needed in the near or far future, the demand planner checks up- coming needs. Warehouse Manager: - It influences this role as it helps prepare for future needs. It affects capacity plan- ning, logistics, as well as playing a key role in communicating with production and knowing what is being produced when. - If sudden rises in production amounts, accurate communication must happen as we need to prepare and ask outsourced warehouses about amounts that they can take. Master Supply planner: - When situations change, communication must be efficient for other teams to be aware of what they need to do next in their role in the supply chain process. - All of these have a hug effect on production, raw material, as well as end ware- housing needs. Warehouse Specialist: - This influences resource planning - Capacity amounts (if more products come into the warehouse than expected) - If we receive less demand than expected, we will overbook employees working within the warehouse. - Checking older batches if IBP / sales/ customer service does not realize (organiz- ing these onto deliveries -> correct communication). Transport and Customs Manager: - First understanding amounts that will be needed for transportation in the future. Whether or not we will use one carrier or multiple ones. - Influences costs as well for transportation. Booking with a certain carrier and not being able to use the booked transportation can cause a fine/penalty. 66 Supply Chain Operations Manager: - Huge impact on service level + stock levels - Determines the stock levels required and thus net working capital. Supply Planner Packaging Material: - Big impact on lead time, transportation - 2–3-month lead time. Business Controller: - Business controller demand, demand planner demand, as well as sales must meet the same scope to fulfill the set amounts. Transport and Customs Coordinator: - We need to check if our allocations with shipping lines are enough, is there enough empty containers, is there enough space on vessels, do we need to make Change of destination and take some containers out in Europe to get them faster to our factory by truck, do we need to order transportation for SPOT purchases? Also, we need to ask suppliers if they can ship something earlier than agreed on the contract or can they on the other hand postpone some of the shipments etc. In ocean freight it is not very easy to make changes to happen fast, but with all the things listed previously we are trying to keep up with forecasts. Also, when tendering ocean freight, we need to give some forecasted amounts of containers, so that we can get as good of a price for ocean freight as possible. So, logistics forecasts influence our work with shipping lines and decision making on how to move certain shipments. 4) Can you share an example of a time when inaccurate demand forecasting had a negative impact on your department’s performance? How did you manage to solve this situation? Demand Planner: - If forecasts are given too late by customers, service level hits happen. Sudden changes in the amounts or format can make forecasting difficult. - Capacity / Batch problems 67 Warehouse Manager: - Sudden campaigns continue and do not stop when they were originally supposed to. Production may continue and therefore there is the need for warehouse work- ers to continue and work overtime. An example: Thursday/Friday overtime needed to be done suddenly when a campaign continued and did not stop on Wednesday when it was supposed to. Communication plays a key role here for all stakeholders involved. Master supply planner: - A good example was when the price of coffee went up. Another business area did not want to forecast the new pricing, whereas another sector did want to. This has caused a mismatch and a never-ending loop when it comes to forecast- ing the correct demand for all raw materials as well as package materials. - This has had a domino effect and has had an impact on seeking where the end products would end up which geographic area) - There were not enough production time / resources to make these corrections. This has an effect in the future to catch up and do what was initially needed. Warehouse Specialist: - Unclear communication / late communication from our warehouse about capac- ity. Outsourced to a 3rd party to operate our warehouse. Sudden changes and the need to move goods elsewhere as our main warehouse are suddenly full. Transport and Customs Manager: - With many green coffee suppliers, we have had stable amounts that are ordered annually. In some situations, we may have peaks in the need of supply, causing forecasting issues. - If the sudden need for more green coffee and/or transportation is needed, prices may be much higher than those whose agreements/purchases were made on time months beforehand. - Supply Chain Operations Manager: 68 - Less was sold than the forecast predicted. Net working capital and return on cap- ital employed declined. Supply Planner Packaging Material: - Forecasts went wrong, leading to not enough packaging material for production. 2–3-month lead time. - Too much was forecasted, packaging material and green coffee used but no sales met. Business Controller: - Supply issues - If differences / process changes are not looked at within different stakeholders of the company, it affects all areas (sales, supply, warehousing, production). Transport and Customs Coordinator: - some new sales and campaigns happened at the last minute, that were not fore- casted beforehand, and it brought a lot of pressure to push shippers and shipping lines to deliver extra. 5) How do external factors like market trends, seasonality, or economic shifts im- pact your demand forecasting and operational planning? Demand Planner: - Seasonal changes play a big role in this role. Seasonal ups and downs, for example the Christmas campaigns coming up. Pricing also fluctuates (green coffee) which has a big impact on the companies’ expenses / amounts that customers are will- ing to buy. Certain constraints with amounts being able to be stored, and there- fore production must be started on time. Stock buffering. Warehouse Manager: - External factors with green coffee supply (comes from across the globe). Not re- ceiving green coffee on time which then is a domino effect. Production cannot produce ready coffee, which then affects forecasting for the warehouse on amounts needed to be stored. - Sudden production machine downtime. This has a huge effect on warehousing as goods need to be stored quickly when the machine is fixed. 69 - Strikes have had a huge effect on operational planning as well in the past. This tends to be at times a seasonal issue. Master Supply Planner: - This has an effect as end consumers do not know what they want. Sudden changes in purchasing trends. Capacity, raw materials, and manpower resources all are affected. - Seasonal changes have a big impact as well. Q4 is very important when it comes to coffee campaigns. Most of the annual sales come from this quarter of the year. Production for these high demands already started a few months prior to the actual sales happening. In this scenario, for example, with the change of products when they are already made, it has a big impact on sales and capacity / ware- housing. - Economic shifts (prices going up), also have a big impact on purchasing power. Even if demand forecasts for sales are done, they can still be changed, and prod- ucts suddenly not be sold. Warehouse Specialist: - Green coffee availability. Projected production planning is set, however, since green coffee comes from across the globe, transport issues may arise. - Sudden changes in costs of raw materials. - Consumer purchasing power changes/fluctuations. - Strikes can affect warehousing as well; do we have enough resources. Not know- ing in time if there is a strike coming or not. Transport and Customs Manager: - Strikes have a big impact on transportation as harbors are closed. - The red sea issue has caused ships to take a different route, causing price changes as well as late arrivals due to longer commute times. - Pandemic times have had a issue on coffee as well. Many people were home dur- ing the pandemic causing larger amounts to be shipped. This has caused freight issues for raw materials. Supply Chain Operations Manager: 70 - Inflation influenced the market as consumer purchasing power - Seasonal trends – more coffee sold at different times of the year. Christmas es- pecially. - Warehouse stock levels are bumped, if possible, strikes are in the scope of the near future. - Coffee pricing is very volatile. - Sudden changes in campaign amounts affecting production and capability to pro- vide to customers on time. Supply Planner Packaging Material: - Holidays, strikes, maintenance breaks - Economic shifts have an impact with pricing. Business Controller: - Consumer purchasing power trends can be highly volatile. - Seasonality differences Transport and Customs Coordinator: - We need to make sure that the amounts that we promise to shipping lines are met and that we don’t “waste” any weekly allocations. If we know that there is a quieter period, we need to communicate that to shipping lines, so we won’t get penalties for not using our agreed space on ocean vessels. And on the other hand, when there is a peak, we need to make sure we have enough allocation agreed and that everything can leave on time. Same with trends. But in ocean freight it is very difficult to make fast changes so 6) What role do you think communication and coordination play in improving the accuracy of forecasts and overall supply chain efficiency? Demand Planner: - Marks a key role in demand planning. Communication must be done in time, to be prepared for production and warehousing. Lead time is a key factor in being able to prepare resources and all. All this plays a key role in mitigating possible service level hits. Demand planner + master planner communication is very im- portant. Notifying in time. 71 Warehouse Manager: - Highly important as the warehouse relies on information coming from the master supply planner, transportation, as well as production. Without the correct infor- mation bottle-neck situations may occur within production. Understanding what goods from production are coming and are going to leave directly, and what will be stored. For example, some goods that come from production are not stored in our warehouse at the main production site but are rather sent directly to our 3PL warehouse partner. Master Supply Planner: - Correct communication within the team (integrated business planning), raw ma- terial demand, package material demand, as well as production demand all must be in the loop if changes are made. Not communicating sudden change in order fulfillment etc. can have a big effect on sales/storage availability etc. - If someone suddenly realizes within the IBP team that within their demand scope something cannot be met, it must be passed onwards outside of the team to- wards sales as well. Warehouse Specialist: - Stakeholders within the company need to communicate on time, to know and see sudden changes on time. - We are at the end of the supply chain, and we are dependent on correct infor- mation from teams working in the previous steps of the supply chain. Transport and Customs Manager: - Communication between different stakeholders within the company is essential. On the other hand, communicating with outside suppliers is very important as well. - Forecasting for machinery is also looked at for transportation in a larger time span, as purchase orders for green coffee are made one year prior to them being shipped. 72 - If changes are made, these need to be communicated. Strikes in different coun- tries or any other issues that may affect delivery. Domino effect in the supply chain. Supply Chain Operations Manager: - Real-time demand data would help forecast future replenishment orders from retailers. - Possible changes need to be communicated to make changes on time within dif- ferent stakeholders working in production, IBP, warehousing etc. - Using a hypothetical approach in forecasting, and not only relying on the num- bers in the system. Supply Planner Packaging Material: - Relying on SAP for forecasts. - Without proper communication knowing what amounts to order would be very hard. Business Controller: - Being proactive and seeking improvements outside of the usual flows can help mitigate risks. - All demand process flows must be communicated to adapt and make changes on time. Transport and Customs Coordinator: - it is everything, without communication nobody knows anything, and it all needs to be re-looked at. As already mentioned, in ocean freight specifically it is impos- sible to make quick changes as the transit times are so long and ships do not sail any faster. So last-minute decisions and bad communication make everything more difficult and usually in ocean freight if something is late, it has a snowball effect and everything from that point on will also be late. So, we always need to be prepared and know what is happening to be able to react. 7) In what ways does demand forecasting affect inventory management, stock re- plenishment, and storage decisions within the warehouse/production of the company? 73 Demand Planner: - This has a huge effect on inventory management, stock replenishment and stor- age decision within warehousing as well as production. Safety stock is highly im- portant. If we see that the demand is going to up, we need to check that we have the correct amount in the safety stock. If demand is lower, we need to check that we don’t have raw materials lying around. Warehouse Manager: - Effects inventory planning for our internal warehouse + external - The correct product demand needs to be known on time. Three key factors come into play; inventory for raw material, packaging material, and finished goods. Master Supply Planner: - It has a big effect. If we have large amounts of production anticipated, not only does filling the warehouse but also distribution days are very hectic. - Some orders may leave late which causes backlogged picking within the ware- house and overlapping of another order that should be picked. - A good example is a big customer sends what they want to purchase in time, to help manage resources for when it is time to pick up and deliver the goods. Warehouse Specialist: - It most affects the need to know do we need to use outsourced warehouses for storing goods. Transport and Customs Manager: - If we do not have any tools used to predict demand for stock replenishment or raw materials, we need to rely on quick reactions. - With correct information in time, and the correct forecasting methods used, we are always aware of amounts needed in, for example, buffer inventory. Supply Chain Operations Manager: - Stock and production amounts are based off forecasts. Supply Planner Packaging Material: - Effects safety stock is updated twice a year to see if changes need to be made / product. 74 Business Controller: - No direct effect for a business controller. Transport and Customs Coordinator: - Same answer as question 5. 8) What role does technology play in enhancing demand forecasting processes in coffee production and warehousing? Demand Planner: - The importance of technology is enhancing. Better forecasting tools always help forecasting. Downsizing from more employees to only one has been done through the help of demand forecasting tools. Production site, product made, customer scope are the ideas that need to be targeted. With the help of technol- ogy, the company has managed to downsize the team to only one person. Warehouse Manager: - The use of SAP ERP as well as excel help enhance processes. - Each container that comes from our green coffee external warehouse is tracked with a GPS that helps timing and tracking of goods to prepare for resource man- agement. Workers can check this to see where the container is at what given time and know the correct time of arrival. Master Supply Planner: - Everything involves forecasting tools as well as ERP software. However, in some cases forecasting tools have sometimes doubled the amount of forecast needed which caused employees to have to manually check whether or not it was correct. - The use of technology plays a key role, however, cannot 100% be always relied on. Warehouse Specialist: - Warehouse management systems (ERP) - WMS – warehouse management systems to track our livestock in our outsourced warehouses 3PL). - Excel for tracking and breaking data sets to understand better what is going on. - Data amounts are so large that without the use of tools we would not be able to track everything and understand different flows and operations used. 75 Transport and Customs Manager: - Technology is not utilized directly for tracking (no transport management system used). - Technology plays a key role in understanding demand forecasting when looking at bigger amounts. Each step of the supply chain is assessed at ease, to pinpoint and seek gaps that can be addressed. - Traceability importance. Supply Chain Operations Manager: - Need to have one predicted forecast that decisions are based on. - Anticipating trends and adjustments. - Using tools to determine when downtime could occur. Supply Planner Packaging Material: - Purchase order monitoring - Claims - Everything is done within the ERP used. - Order management + stock for case company 1 as well as 2. Business Controller: - Right tools to implement/ and ease flows - Helps bump of quality levels. - Seeing demand with the use of different tools helps see the overall scope. Doing everything manually takes time and resources. Transport and Customs Coordinator: - better tracking tools for our shipments make forecasting easier, and a good ERP system also. Technology is very helpful, and I would like to see even more tools developed for this. 9) What are the key factors influencing forecasting accuracy in the coffee industry (influence your department)? Demand Planner: 76 - Being able to predict future campaigns and how much should be made based on the buying power of consumers. Being able to forecast the correct amounts for a given week. Warehouse Manager: - Being aware of future campaigns from our biggest customers in Finland and Bal- tics. Master Supply Planner: - The right communication for timing for the campaigns. Timing for which week the production will start, as well as what week will they be delivered. - Understanding which products are going to move out of the warehouse at which time. Analyzing risks from previous trends to seek possible gaps that can be mit- igated. Warehouse Specialist: - Correct information and that the demand forecasts are done correctly. - Being too dependent on previous demand peaks and seasonal changes. Under- standing that they may change and being prepared for a different time ahead. Transport and Customs Manager: - When campaigns are created for a certain product, adapting to change and for example the change of a product can have a huge impact on production and warehousing. - SKU level can be changed; however, end consumers may not understand that the green coffee used for another product may not be the same. - Understanding and seeing consumer market trends. Supply Chain Operations Manager: - Pricing campaigns - How competitor campaigns influence sales. Supply Planner Packaging Material: - Communication - Correct amounts in SAP (which come from other supply planners / sales). Business Controller: 77 - Teamwork to meet required amounts for demand - Correct communication on time within the company as well as retail customers to understand the buying amounts of others as well as what we can produce. Transport and Customs Coordinator: - communication about new campaigns, trends and seasonality, analyzing and considering all the possible risks involved, making sure that all parties are aware of any changes and that everyone is on the same page. 10) How do you think collaboration between stakeholders (production, warehous- ing, sourcing) improves the accuracy of operational tasks in the coffee supply chain? Demand Planner: - Highly important, when we know the demand, we can communicate with sourc- ing and production the amounts needed to start production. Warehouse Manager: - Weekly/monthly catch ups or meetings with other stakeholders from production, indirect sourcing, quality, technology department help understand current ongo- ing problems or future trends. - Communication playing again a key role. Master Supply Planner: - Everyone works together within the company to meet customer requirements. Adapting to changes when something is happening in another team to possibly make changes elsewhere. Communication is a big plus. Warehouse Specialist: - Weekly and monthly meetings to see changes. Correct communication. Too much communication does not exist. - Without correct communication the supply chain loop may have gaps causing problems in the end warehousing. Transport and Customs Manager: - Helps ease and plan resources and need for employees working in the warehouse. 78 - Understanding gaps that can be found when communicating on a weekly or monthly basis. - Communicating on time for other stakeholders being able to make their own nec- essary changes if needed. Supply Chain Operations Manager: - Highly important. Supply Planner Packaging Material: - Highly important if everyone understands forecasts as well as inventory is correct. It is very important to communicate with our outsourced warehouse workers to count amounts physically at times to see that the stock is the same as in the ERP system. Business Controller: - Adapting to change - Overseeing possible issues Transport and Customs Coordinator: - very important, we have had many incidents where communication has not worked as good as it should have, and it makes working more difficult and stress- ful. Collaborations not only makes things go smoother, but it also causes less stress and can save money when we are prepared to do certain things i/o just trying to find “whatever” solutions. 11) How do you think artificial intelligence will help demand forecasting and oper- ational tasks soon within production and warehousing? Demand Planner: - When AI demand is enhanced enough it can be utilized to help see what will be made, when, as well as what volumes. Capacity is a key factor. Within production and warehousing AI is not affected now. If we were to get enough information from different steps (involving capacity) AI could be used to help demand. There are so many steps involving maximum capacity and different options of produc- tion, which is very hard to implement now. This is something that can be utilized 79 possibly in the near or far future. Implementing the correct information towards AI and its importance. Warehouse Manager: - Artificial intelligence can help calculate future needs in a short-term perspective or even long term - Complicated data sets can be used to seek better options for storage concerns. - Automated warehousing, which is something that has already been spoken about, can help minimize manpower as well as efficiency. Master Supply Planner: - It will help if/ and when people understand how artificial intelligence works. The correct prompt. Consumer buying power, sales, and all will increase. Warehouse Specialist: - It will help if we have enough data from the past to have the correct input for AI to give future predictions. - Having someone check that the AI forecasts seem correct and not relying on that everything is correct. - AI will have a big impact in the future on demand forecasting. Warehouse capac- ity, automating warehousing for better picking and accuracy. Transport and Customs Manager: - Using artificial intelligence tools when considering transportation of ship freight (looking at possible strikes that may happen in the near or far future, weather forecasts etc.). - Even if AI is used, someone should overlook this and check that it has not pre- dicted wrong. As we gain more information in the future, it can be better relied on. - AI will be utilized even more in the future when enough data has been gathered on consumer trends. Coffee sales depend on consumer buying power, and that is something that AI can most likely predict even better than a subject matter expert. Supply Chain Operations Manager: 80 - Understanding campaigns (for example ways of marketing) - Forecast accuracy for campaigns. Supply Planner Packaging Material: - Will help seek trends and see forecast history - Handling basic data and not causing someone to do it manually Business Controller: - Will help in the future for predicting demand, however, someone needs to check that the data is correct. - May not take into consideration all key factors that a human can see. Will help with downsizing resources. Transport and Customs Coordinator: - I am sure it will somehow. Maybe it will be able to better find all the risk factors. It may also make suggestions that we haven’t thought about. 12) Looking forward, how do you see the role of demand forecasting evolving in the supply chain? What future trends or innovations do you think will impact how forecasts are created and used? Demand Planner: - The future need for a demand planner is something nobody knows. Will artificial intelligence be able to predict all of this. Since green coffee is not something that can be always predicted, manual tasks may still be needed in the future as AI cannot always predict these. Warehouse Manager: - Traceability for raw materials from the original destination will advance, which help prepare for preparing production as well as warehousing needs. - Possible Transport Management Systems that can pinpoint and help routing. - Artificial intelligence can see different trends in markets that someone may not notice and mitigate a possible risk. Master Supply Planner: - The importance of demand forecasting will increase in the future. 81 - Everything is happening faster, and therefore the demand forecast must be more accurate (this will be enhanced with AI as it will look at many factors that come into play). - Warehouse services as well as rent will go up, and downsizing within warehouse capacity will happen. Therefore, accuracy in demand forecasting must be more accurate to make these happen correctly. Warehouse Specialist: - Automating demand forecasting with different tools will help predict stock re- plenishment, stock transfer orders to outsourced warehouses etc. - More accurate than what it is now, even though different tools are being used as well now. - Being able to understand from a budgeting perspective will be more accurate. Knowing what goods will be cheaper to store in different warehouses using AI and demand tools. - Not needing so much work power to make decisions and seek different options. Being able to get different options from demand forecasting tools or AI. Transport and Customs Manager: - More accurate predictions - Downsizing warehouse capacity and being able to utilize a ready product in and out cycle much faster. Supply Chain Operations Manager: - Real life integration with those buying can help understand forecasts. - Merging different roles as AI will help with these (currently done manually) - More of a monitoring system approach - Automation Supply Planner Packaging Material: - Lots of variables need to be taken into consideration from who makes the prod- uct all the way to end production. - Hard to monitor, most likely some work power will be needed in the future, how- ever AI can help automate some parts. 82 Business Controller: - Economical shifts as well as possible constraints may not ease how the supply chain works in the future, causing a possible change in different flows. Transport and Customs Coordinator: - I think it is and will be very important always. Maybe in the future it will be more automated / more tools will be used to help people. Difficult to say. 83 7.2 Sample Data