RESEARCH ARTICLE
Measuring user interactions with websites: A
comparison of two industry standard
analytics approaches using data of 86
websites
Bernard J. JansenID1☯*, Soon-gyo Jung1☯, Joni Salminen1,2,3☯
1 Qatar Computing Research Institute, Hamid Bin Khalifa University, Doha, Qatar, 2 Turku School of
Economics, University of Turku, Turku, Finland, 3 School of Marketing and Communication, University of
Vaasa, Vaasa, Finland
☯ These authors contributed equally to this work.
* jjansen@acm.org
Abstract
This research compares four standard analytics metrics from Google Analytics with Similar-
Web using one year’s average monthly data for 86 websites from 26 countries and 19 indus-
try verticals. The results show statistically significant differences between the two services
for total visits, unique visitors, bounce rates, and average session duration. Using Google
Analytics as the baseline, SimilarWeb average values were 19.4% lower for total visits,
38.7% lower for unique visitors, 25.2% higher for bounce rate, and 56.2% higher for session
duration. The website rankings between SimilarWeb and Google Analytics for all metrics
are significantly correlated, especially for total visits and unique visitors. The accuracy/inac-
curacy of the metrics from both services is discussed from the vantage of the data collection
methods employed. In the absence of a gold standard, combining the two services is a rea-
sonable approach, with Google Analytics for onsite and SimilarWeb for network metrics.
Finally, the differences between SimilarWeb and Google Analytics measures are system-
atic, so with Google Analytics metrics from a known site, one can reasonably generate the
Google Analytics metrics for related sites based on the SimilarWeb values. The implications
are that SimilarWeb provides conservative analytics in terms of visits and visitors relative to
those of Google Analytics, and both tools can be utilized in a complementary fashion in situ-
ations where site analytics is not available for competitive intelligence and benchmarking
analysis.
Introduction
Web analytics is the collection, measurement, analysis, and reporting of digital data to enhance
insights concerning the behavior of website visitors [1]. Web analytics is a critical component
of business intelligence, competitive analysis, website benchmarking, online advertising,
online marketing, and digital marketing [2] as business decisions are made based on website
PLOS ONE
PLOS ONE | https://doi.org/10.1371/journal.pone.0268212 May 27, 2022 1 / 27
a1111111111
a1111111111
a1111111111
a1111111111
a1111111111
OPEN ACCESS
Citation: Jansen BJ, Jung S-g, Salminen J (2022)
Measuring user interactions with websites: A
comparison of two industry standard analytics
approaches using data of 86 websites. PLoS ONE
17(5): e0268212. https://doi.org/10.1371/journal.
pone.0268212
Editor: Hussein Suleman, University of Cape Town,
SOUTH AFRICA
Received: February 2, 2021
Accepted: April 25, 2022
Published: May 27, 2022
Peer Review History: PLOS recognizes the
benefits of transparency in the peer review
process; therefore, we enable the publication of
all of the content of peer review and author
responses alongside final, published articles. The
editorial history of this article is available here:
https://doi.org/10.1371/journal.pone.0268212
Copyright: © 2022 Jansen et al. This is an open
access article distributed under the terms of the
Creative Commons Attribution License, which
permits unrestricted use, distribution, and
reproduction in any medium, provided the original
author and source are credited.
Data Availability Statement: The data underlying
the results presented in the study are available
from SimilarWeb (https://www.similarweb.com/).
traffic measures obtained from website analytics services. Organizations monitor their sites’
incoming and outgoing traffic to identify popular pages, determine user interests, and stay
abreast of emerging trends [3]. There are various ways to monitor this traffic, and the gathered
data is used for re-structuring sites, highlighting security problems, indicating bandwidth
issues, assessing organizational key performance indicators (KPIs), and obtaining societal
insights [4].
Approaches to collecting website analytics data can be grouped by the focus of data collec-
tion efforts, resulting in the emergence of three general methodologies, namely: (a) user-cen-
tric, (b) site-centric, and (c) network-centric. The central traits of each are as follows.
• User-centric: Web analytics data is gathered via a panel of users, which is tracked by soft-
ware installed on users’ computers, such as a plugin for a web browser [5–8]. For example,
when users install an extension to their browser, they approve in the license agreement that
the data on the websites they visit will be processed and analyzed. The primary advantage
here is that the user-centric approach does not rely on cookies or tags (i.e., snippets of infor-
mation placed by a server to a user’s web browser in order to keep track of the user) but on
direct observation. An additional advantage is comparing web analytics data across multiple
websites. The challenge is recruiting and incentivizing a sufficiently large user panel that is a
representative sample of the online population—due to this challenge, only a few companies
have recruited sizeable user panels (e.g., Alexa). Another disadvantage may be the issue of
privacy since many users are not willing to share information on every website that they
visit, so some users may make efforts to mask their actual online actions from the tracking
plugin.
• Site-centric: Web analytics is gathered via software on a specific website [9–16]. Most web-
sites use a site-centric approach for analytics data gathering, typically employing cookies
and/or tagging pages on the website (e.g., Google Analytics, Adobe Analytics). The primary
advantage of this approach lies in counting events and actions (e.g., pages viewed, times
accessed), which is relatively straightforward. Another advantage is that users do not need to
install specific software beyond the browser. However, there are disadvantages. First, site-
centric software focuses on cookies/tags, so these counts may not reflect actual people (i.e.,
the measures are of the cookies and tags) or people’s actual actions on the website. Instead,
site-centric approaches measure the number of cookies dropped or tags fired as proxies for
people or interactions. Second, this approach is susceptible to bots (i.e., autonomous pro-
grams that pretend to be real users) and other forms of analytics inflation tactics, such as
click fraud [17]. Finally, the site-centric analytics usually represent just one website and are
only accessible to the owner of that website, making the site-centric approach not widely
available for business intelligence, marketing, advertising, or other tasks requiring web ana-
lytics data from a large number of sites.
• Network-centric: Web analytics is gathered via observing and collecting traffic in the net-
work [18, 19]. There are various techniques for network-centric web analytics data gather-
ing, with the most common being data purchased or acquired directly from Internet service
providers (ISPs). However, other data gathering methods include leveraging search traffic,
search engine rankings, paid search, and backlinks [20, 21]. The main advantage of the net-
work-centric approach is that one can relatively easily collect analytics concerning a large
number of websites. Also, the setup is comparatively easy, as neither users nor websites are
required to install any software. The major disadvantage is that there is no information
about the onsite actions of the users. A second disadvantage is that major ISPs do not freely
share their data, so acquiring it can be expensive. However, companies can acquire other
PLOS ONE Measuring user interactions with websites
PLOS ONE | https://doi.org/10.1371/journal.pone.0268212 May 27, 2022 2 / 27
The authors had no special access privileges to the
data others would not have.
Funding: The author(s) received no specific
funding for this work.
Competing interests: The authors have declared
that no competing interests exist.
network-centric data more reasonably (i.e., SpyFu, SEMRush; two common industry tools
for search marketing), albeit requiring substantial computational, programming, and storage
resources.
Of course, one can use a combination of these methods [22], but these are three
general approaches, with much academic research leveraging one or more of these methods
[23–26]. See Table 1 for a summary of the advantages, disadvantages, and examples of
implementations.
While site-centric web analytics tools, such as Google Analytics, can provide results for
one’s own website, there is often a need to compare with other websites, though Google Ana-
lytics does provide some limited benchmarking reports by industry (https://support.google.
com/analytics/answer/6086666). Therefore, competitive benchmarking services, such as Simi-
larWeb, have become essential for web analytics in the business intelligence area [27]. These
analysis services provide computational web analytics results for one or more websites, a criti-
cally needed capability for competitive research and analysis [28]. These website analytics ser-
vices allow benchmarking of web analytics measures and metrics among multiple websites.
Website analytics services are essential for a variety of reasons, including competitive analysis,
advertising, marketing, domain purchasing, programmatic media buying [29–35], and firm
acquisition [36], along with the use of website analytics services in academic research [37, 38].
They are also valuable for accessing the external view of one’s own website (i.e., what others
who do not have access to site-centric analytics data see). These website analytics services
return a variety of metrics depending on the platform. However, there are questions concern-
ing the accuracy and reliability of both types of analytics platforms, affecting billions of dollars
in online advertising, firm acquisition, and research. As such, there is a critical need to assess
these tools and the validity of the reported metrics.
In this research, we compare web analytics statistics from Google Analytics (the industry-
standard website analytics platform at the time of the study) and SimilarWeb (the industry-
standard traffic analytics platform at the time of the study) using four core web analytics met-
rics (i.e., total visits, unique visitors, bounce rate, and average session duration) averaged
monthly over 12 months for 86 websites. We select SimilarWeb due to the scope of its data col-
lection, reportedly one billion daily digital signals, two terabytes of daily analyzed data, more
than two hundred data scientists employed, and more than ten thousand daily traffic reports
generated, with reporting features better or as good than other services [39] at the time of the
study. As such, SimilarWeb represents state-of-the-art in the online competitive analytics area.
We leave the investigation of others services besides Google Analytics and SimilarWeb to
other research. We conduct statistical analysis along several fronts reporting both exploratory
Table 1. Comparison of user, site, and network-centric approaches to web analytics data collection showing advantages, disadvantages, and examples of each
approach at the time of the study.
Approach Advantages Disadvantages Examples
User-centric • Focus on people
• Compare across websites; so can use for business
intelligence
• Creating a representative user panel is challenging
• User computer software must be installed
• Alexa
• ComScore
Site-Centric • No special user software to install
• Wide range of analytics for a specific site
• Site software must be installed
• Focus on cookies and tags, not real people
• Access is limited to the website owner; cannot use for business intelligence
among multiple sites
• Google
Analytics
• Adobe
Analytics
• IBM Analytics
Network-
Centric
• Data collection is straightforward
• No special software to install for users or sites
• Compare across websites; can use for business
intelligence
• Data can be challenging to obtain
• Limited onsite analytics; generally only between sites data
• Hitwise
• SEMRush
• SpyFu
https://doi.org/10.1371/journal.pone.0268212.t001
PLOS ONE Measuring user interactions with websites
PLOS ONE | https://doi.org/10.1371/journal.pone.0268212 May 27, 2022 3 / 27
and statistical results. We then tease apart the nuanced differences in the metrics and possible
sources of error [40] and present the theoretical and the practical implications of this research.
The techniques employed by Google Analytics are similar to techniques employed by other
analytics platforms, such as Adobe Analytics, IBM Analytics, and Piwik Analytics. The tech-
niques used by SimilarWeb are similar to the techniques of other website analytics services,
such as Alexa, comScore, SEMRush, Ahrefs, and Hitwise, in the employment of user, site, and/
or network data collection. So, the results of this research apply to a wide range of analytics
tasks, most notably in the website domain, providing an enhanced understanding of the data
underlying competitive intelligence and the use of such analytics platforms. Moreover, the
metrics reviewed are commonly used in many industries employing online analytics, such as
advertising, online content creation, and e-commerce. Therefore, the findings are impactful
for several domains.
Review of literature
Web analytics services have been employed in research and used by researchers for an array of
inquiries and topics. These areas include, among others, online gaming [41], social media and
multi-channel online marketing [42, 43], online community shopping [44], online purchase
predictions [45, 46], online research methods [47], social science [48, 49], and user-generated
content on social media [50–54]. These services have also been used in research concerning
online interests in specific topics [55–57]–including online branding in social media [58, 59],
online purchasing [60], and mobile application usage [61]. They have also been used in studies
about website trust and privacy [62–66], website design [37, 67–69], and website popularity
and ranking [42, 44, 70–77]–for a variety of areas. These prior studies indicate that analytics
tools are widely used in peer-reviewed academic research and relied on for various metrics.
However, to our knowledge, none of the prior research studies examined the accuracy of these
website analytics services prior to employment.
Academic research on this area of analytics evaluation is limited. Lo and Sedhain [78] evalu-
ate six websites lists, including the ranked list from Alexa (the only service employed in the
study that is still active, as of the date of this research). The researchers examined the top 100
websites and compared the rankings among the lists. They concluded that the ranking among
the lists differed. This difference is not surprising given that the methodologies used to create
the study lists varied in terms of website traffic, number of backlinks, and opinions of human
judges. Vaughan and Yang [79] use organizations from the United States (U.S.) and China and
collect web traffic data for these sites from Alexa Internet, Google Trends for Websites, and
Compete (Alexa is the only service still active from the study, as of the date of this research).
The researchers did not evaluate the traffic services but instead reported correlations between
web traffic data and measures of academic quality for universities. In a ComScore study,
Napoli, Lavrakas, and Callegaro [80] present some of the challenges and issues with the user-
centric analytics approach, namely that the results often do not align with site-centric mea-
sures. The researchers attribute the discrepancies to the sampling of the user panels. Scheitle
and fellow researchers [19] examine several websites’ rankings, including Alexa but not Simi-
larWeb, investigating similarity, stability, representativeness, responsiveness, and benignness
in the cybersecurity domain, but they do not report actual analytics numbers. The researchers
report that the ranked lists are unstable and open to manipulation. Pochat and colleagues [18]
extend this research by introducing a list that is less susceptible to rank manipulation.
While few academic studies have examined analytics services, fewer have evaluated the
actual analytics numbers; instead, they focus on the more easily accessible (and usually free)
ranked lists. Studies are even rarer still on the performance of SimilarWeb, despite its standing
PLOS ONE Measuring user interactions with websites
PLOS ONE | https://doi.org/10.1371/journal.pone.0268212 May 27, 2022 4 / 27
and reputation as an industry leader. Scheitle and colleagues [19] attribute this absence to
SimilarWeb charging for its service, although the researchers do not investigate this conjec-
ture. Regardless of the reason, the only academic study that we are aware of as of the date of
this research that explicitly examines traffic numbers, including SimilarWeb, is Prantl and
Prantl [24]. This study compares rankings among Alexa, SimilarWeb, and NetMonitor [80]
for a set of websites in the Czech Republic, using NetMonitor as the baseline. The research
only reports the traffic comparison between SimilarWeb and NetMonitor. The researchers,
unfortunately, provide neither detailed exploratory analysis nor statistical analysis of the
analytics comparison. Also, NetMonitor uses a combination of site and user-centric measures,
so it is unclear how the traffic metrics are calculated. The researchers [24] report that Similar-
Web over reports traffic compared to NetMonitor. They also note that SimilarWeb traffic
results are +/- 30% compared to NetMonitor traffic measurements for 49% of the 487
websites.
Several industry reports have also compared site analytics, usually using Google Analytics,
with the analytics reported by other services. Some of these reports [81–83] show website ana-
lytics services, notably SimilarWeb, reportedly underestimating traffic, as much as 30% to
50%, while other reports (84–88] claim SimilarWeb overestimates traffic, from 11% for large
websites to nearly 90% for small ones [84]. SimilarWeb itself states that reported values among
analytics services will vary +/- 20%. However, a trend is that SimilarWeb [85, 86] consistently
ranks as the best or one of the best analytics services in the industry [87, 88], as noted by sev-
eral industry practitioners [30, 32, 33, 89–91]. SimilarWeb consistently outperforms other ser-
vices [92], with reported performance better sometimes in the double digits [93]. Even when
the reported analytics numbers are off, the SimilarWeb results usually correlate with the base-
line site traffic trends. The correlation is also positive relative to overall accuracy among sites
[93].
Although providing insights into the area, there are potential issues regarding relying on
industry reports, including possible questions on data appropriateness, lack of explicitly
defined methods of analysis, and conflicts of interest (as some of these studies are performed
by potential competitors of SimilarWeb). Also, some of these studies employ a small number
of data points [81, 82, 94], making statistical analysis challenging. Other studies have a short
temporal span [83, 88, 93], as there can be significant traffic fluctuations for sites depending
on the time of year, or mainly high-traffic websites [83], which are easier to calculate. Finally,
some reports have imprecise metric reporting [83, 84, 92, 95], raising doubt on the results, or a
limited set of metrics [81, 83, 95] not central to analytics insights. Because of these potential
issues, there is a critical need for a rigorous academic analysis of website analytics services to
supplement these industry reports.
Given the substantial use of analytics services in academic research and their widespread
use in the practitioner communities, there is a notable lack of research examining the accuracy
of these services. Determining their accuracy is critical, given the extensive reliance on analyt-
ics numbers across many domains of research and practice [96]. However, due to the absence
of academic studies in the area, several unanswered questions remain, including: How accurate
are these analytics services? How do they compare with other analytics methods? Are these ana-
lytics tools better (or worse) at measuring specific analytics metrics than other methods? Are the
reported metrics valid? These are essential questions that need addressing for critical evaluation
of research findings and business decisions that rely on these services. Although the questions
are conceptually straightforward, they are surprisingly difficult to evaluate in practice. This dif-
ficulty, especially in terms of data collection, may be a compounding factor for the dearth of
academic research in the area.
PLOS ONE Measuring user interactions with websites
PLOS ONE | https://doi.org/10.1371/journal.pone.0268212 May 27, 2022 5 / 27
Research questions
Our research objective is to compare and contrast the reported analytics measurements between
SimilarWeb and Google Analytics in support of the broader goal of comparing these two
approaches for measuring analytics and evaluating their accuracy. To investigate this research
objective, we focus on four core web analytics metrics–total visits, unique visitors, bounce rate,
and average session duration–which we define in the methods section. Although there is a
lengthy list of possible metrics for investigation, these four metrics are central to addressing
online behavioral user measurements, including frequency, reach, engagement, and duration,
respectively. We acknowledge that there may be some conceptual overlap among these met-
rics. For example, bounce rates are sessions with an indeterminate duration that may indicate
a lack of engagement, but average session duration also provides insights into user engage-
ment. Nevertheless, these four metrics are central to the web analytics analysis of nearly any
single website or set of websites; therefore, they are worthy of investigation. In the interest of
space and impact of findings, we focus on these four metrics, leaving other metrics for future
research.
Given that Google Analytics uses site-centric website data and SimilarWeb employs a trian-
gulation of datasets and techniques, we would reasonably expect values would differ between
the two. However, is it currently unknown how much they differ, which is most accurate, or if
the results are correlated. Therefore, because Google Analytics is, at the time of the study, the
de facto industry standard for websites, we use Google Analytics measurements as the baseline
for this research. Our hypotheses (H) are:
• H1: SimilarWeb measurement of total visits to websites differ from those reported by Google
Analytics.
• H2: SimilarWeb measurement of unique visitors to websites differ from those reported by
Google Analytics.
• H3: SimilarWeb measurement of bounce rates for websites differ from those reported by
Google Analytics.
• H4: SimilarWeb measurement of average session durations for websites differ from those
reported by Google Analytics.
We investigate these hypotheses using the following methodology.
Material and methods
Our data collection platforms are Google Analytics and SimilarWeb. Each service is explained
in the following subsections.
Google analytics
Google Analytics is a site-centric web analytics platform and, at the time of the study, is the
most popular site analytics tool in use [97]–that is, it is the market leader. Google Analytics
tracks and reports website analytics for a specific website. This tracking by Google Analytics is
accomplished via cookies and tags [98]; a tag is a snippet of JavaScript code added to the indi-
vidual pages. The tags are executed in the JavaScript-enabled browsers of the website visitors.
Once executed, the tag sends the visit data to a data server and sets a first-party cookie on
cookie-enabled browsers on visitors’ computers. The tag must be on a page on the site for Goo-
gle Analytics to track the web analytics data for that page.
PLOS ONE Measuring user interactions with websites
PLOS ONE | https://doi.org/10.1371/journal.pone.0268212 May 27, 2022 6 / 27
Concerning the data collection, analysis, and reporting algorithms of Google Analytics,
they are proprietary. However, enough is known to validate their employment as being indus-
try standard and state-of-the-art. The techniques of cookies and the general process of tagging
are well-known, although there may be some nuances in implementation. Google Analytics
employs statistical data sampling techniques [99], so the values in these cases may not be the
result of the complete data analysis for some reports. However, the general overview of the
data sampling approach is presented in reasonable detail [29], and the described subsampling
is an industry standard methodology [100].
SimilarWeb
SimilarWeb [85, 86] is a service providing web analytics data for one or multiple websites.
SimilarWeb uses a mix of user, site, and network-centric data collection approaches to triangu-
late data [39, 101], reportedly collecting and analyzing billions of data points per day [22].
SimilarWeb’s philosophical approach is that each method has strengths and weaknesses, and
the best practice is triangulating multiple algorithms and data sources [39], a respected
approach in data collection and analysis.
Regarding the data collection, analysis, and reporting algorithms of SimilarWeb, they are
proprietary, but again, enough is known to validate the general implementation as state-of-
the-art. The SimilarWeb foundational principle of triangulating user, site, and network-centric
data collection data [39, 101] is academically sound, with triangulating data and methods used
and advocated widely by scholars [5, 102]. SimilarWeb data collection, analysis, and reporting
methodology are outlined in reasonable detail [22], although, like Google Analytics, the pro-
prietary specifics are not provided. However, from the ample documentation that is available
[22, 39, 86, 103–105], the general approach is to collect data from three primary sources, which
are: (a) a reportedly 400 million worldwide user panel [103] at the time of the study, (b) spe-
cific website analytics tracking [39], and (c) ISP and other traffic data [39]. These sources are
supplemented with publicly available datasets (e.g., population statistics). Each of these data-
sets will overlap (i.e., the web analytics data from one collection method will also appear in one
or both of the other collection methods). With the collected data augmented with publicly
available data [39], SimilarWeb uses statistical techniques and ensemble machine learning
approaches to generate web analytics results. These analytics can then be compared to the
overlapped data to make algorithmic adjustments to the predictions. This is a more complex
approach relative to Google Analytics; however, SimilarWeb’s scope of multiple websites also
requires a more complicated approach. In sum, the general techniques employed by Similar-
Web are standard methodologies [101, 106, 107], academically sound, and industry standard
state-of-the-art.
Data collection procedure
For our analysis, we identify a set of websites with analytics by SimilarWeb and having their
Google Analytics accessible by SimilarWeb [104, 108], thereby making their Google Analytics
values available. If a website has a Google Analytics associated, SimilarWeb, using the paid ver-
sion, offers the option of reporting either the SimilarWeb or the Google Analytics numbers for
these websites. For this access, the website owner grants SimilarWeb access to the website’s
Google Analytics account, so the data pull is direct. We verified this process with a website not
employed in the study, encountering no issues with either access or reported data. This feature
allows us to compare the SimilarWeb and the Google Analytics numbers for our identified
web analytics metrics of total visits, unique visitors, bounce rates, and average session
duration.
PLOS ONE Measuring user interactions with websites
PLOS ONE | https://doi.org/10.1371/journal.pone.0268212 May 27, 2022 7 / 27
We employ the Majestic Million [108] to identify our pool of possible websites. The Majes-
tic Million list of websites is creative commons licensed and derives from Majestic’s web
crawler. The Majestic Million list ranks sites by the number of /24 IPv4-subnets linking to that
site, used as a proxy for website popularity. Using this large, open-licensed, and readily avail-
able list as the seed listing, we started at the top, submitted the link to the SimilarWeb applica-
tion program interface (API), and checked whether SimilarWeb provided analytics or if the
website associated its Google Analytics to the SimilarWeb service. We included it as a candi-
date website for our research if it had both SimilarWeb and Google Analytics metrics. If not,
the website was excluded. We then proceeded to the following website on the list and repeated
the submission and verification process.
We continued these steps until we identified 91 websites. There were five websites where
Google Analytics and SimilarWeb values differed by orders of magnitude. As there seemed to
be no discernible patterns among these five websites upon our examination, we excluded them
as outliers and reserved them as candidates for future study. This action left us with 86 websites
for analysis. We concluded that this was more than satisfactory for our research, as the number
is adequate for statistical analysis [109].
We have determined not to make the specific links publicly available for the privacy of the
companies’ websites and given that these web analytics comparisons are a paid business prod-
uct of SimilarWeb. However, we outline our methodology in detail so that those interested can
recreate our research. Also, we provide the web analytics and related data concerning the web-
sites (excluding website name and website link) in S1 File.
Data analysis
We employ paired t-tests for our analysis. The paired t-test compares two means from the
same population to determine whether or not there is a statistical difference. As the paired t-
test is for normally distributed populations, we conduct the Shapiro-Wilk test for visits, unique
visits, bounce rate, and average session duration for both platforms to test for normality. As
expected, the Shapiro-Wilk tests showed a significant departure from the normality for all vari-
ables. Therefore, we transformed our data to a normal distribution via the Box-Cox transfor-
mation [110] using the log-transformation function, log(variable). We then again conducted
the Shapiro-Wilk test; the effect sizes of non-normality were very small, small, or medium,
indicating the magnitude of the difference between the sample and normal distribution.
Therefore, the data is successfully normalized for our purposes, though a bit of skewness exists,
as the data is weighted toward the center of the analytics numbers using the log transforma-
tion, as shown for visits in Fig 1.
Despite the existing skewness, previous work shows that a method such as the paired t-test
is robust in these cases [111, 112]. The transformation ensured that our statistical approach is
Fig 1. Histogram of normalized Google Analytics and SimilarWeb visits data. Effect sizes Are Very Small and Small Respectively, Indicating
the Difference Between the Sample Distribution and the Normal Distribution is Very Small/Small.
https://doi.org/10.1371/journal.pone.0268212.g001
PLOS ONE Measuring user interactions with websites
PLOS ONE | https://doi.org/10.1371/journal.pone.0268212 May 27, 2022 8 / 27
valid for the dataset’s distributions. We then execute the paired t-test on four groups to test the
differences between the means of total visits, unique visitors, bounce rates, and average session
duration on the transformed values.
Further, we employ the Pearson correlation test, which measures the strength of a linear
relationship between two variables, using the normalized values for the metrics under evalua-
tion. This correlation analysis informs us how the two analytics services rank the websites rela-
tive to each other for a given metric, regardless of the agreement on the absolute values. These
analytics services are often employed in site rankings, which is a common task in many com-
petitive intelligence endeavors and used in many industry verticals, so such correlation analysis
is insightful for using the two services in various domains.
Using the SimilarWeb API, we collect the reported values for total visits, unique visitors,
bounce rate, and average session duration for each month over 12 months (September 1, 2019,
through August 31, 2020, inclusive) for each of the 86 websites on our list. We then average the
monthly values for each metric for each platform to obtain the values that we use in our analy-
sis. We use the monthly average to mitigate any specific monthly fluctuation. For example,
some websites have seasonal fluctuations in analytics. Some websites may experience outages
during specific months or denial of service attacks. Using the monthly average over 12 months
helps mitigate the possible short-term variations.
Our four measures, total visits, unique visitors, bounce rate, and average session duration,
are considered core metrics in the domain of web analytics [1, 113, 114]. A metric is typically a
number, such as a count or a percentage. However, measuring or calculating these metrics
may vary by platform or service; therefore, it is crucial to understand these differences. Addi-
tionally, the conceptual understanding of these metrics may differ from the specific ability of a
method for tracking in implementation. Table 2 presents an overview of these metrics.
Results
Exploratory results
Our 86 websites represent companies based in 26 countries, as shown in Table 3. We used the
country classifications provided by SimilarWeb, and we verified the classifications based on
our assessment of the websites and links.
The 86 organizational websites are from the following 19 industry verticals, as shown in
Table 4. We used the industry classifications provided by SimilarWeb [115, 116], and we veri-
fied the classifications based on our assessment of the websites and company background
material provided.
The types of the 86 organizational websites are shown in Table 5. We used the site type clas-
sifications provided by SimilarWeb, and we verified the classification based on our assessment
of the website content and features. Content sites are websites that provide content as their pri-
mary function. Transactional websites are sites that are primarily selling a product. ‘Other’
refers to those websites that do not fit into the other two categories.
H1: Measurements of total visits differ
A paired t-test was conducted to compare the number of total visits reported by Google Ana-
lytics and SimilarWeb. There was a significant difference in the reported number of total visits
for Google Analytics (M = 6.82, SD = 0.31) and SimilarWeb (M = 6.66, SD = 0.29); t(85) =
6.43, p< 0.01. These results indicate a difference in the number of total visits between the two
approaches. Specifically, our results show that SimilarWeb’s reported number of total visits is
statistically lower than the values reported by Google Analytics. Therefore, H1 is fully
PLOS ONE Measuring user interactions with websites
PLOS ONE | https://doi.org/10.1371/journal.pone.0268212 May 27, 2022 9 / 27
supported: SimilarWeb’s measurements of total visits to websites differ from those
reported by Google Analytics.
The number of total visits for all 86 websites was 1,703.5 million (max = 292.5 million;
min = 1,998, med = 7.8 million), as reported by Google Analytics, and 1,060.1 million
(max = 140.8 million; min = 4,443; med = 5.9 million), as reported by SimilarWeb. Using the
total aggregate visits for all 86 websites using Google Analytics as the baseline, SimilarWeb
Table 2. Comparison of definitions of total visits, unique visitors, bounce rate, and session duration conceptually and for the two analytics platforms: Google Ana-
lytics and SimilarWeb.
Definition of: Total Visits Unique Visitors Bounced Rate Average Session Duration
Conceptually Sum of times that all people go to a
website during a measurement
period.
A measure of frequency.
Sum of actual people who have
visited a website at least once
during a period.
A measure of reach.
A bounced visit is the act of a person
immediately leaving a website with
no interaction.
A measure of engagement.
The average length of time that
visitors are on the website.
A measure of duration.
Practically Sum of times at least one page of a
website has been loaded into a
browser during a measurement
period.
Sum of distinct tracking measures
requesting pages from a website
during a given period determined
by a method such as cookie, tag,
or plugin.
Ratio of single-page visits divided by
all visits to a website during a given
period (i.e., single page visits divided
by all visits)
Total duration of all sessions
divided by the number of sessions
Google
Analytics
Sum of single visits to a website
consisting of one or more pageviews
during a measurement period.
The default visit timeout is 30
minutes, meaning that if there is not
activity for this visit on the website
for more than 30 minutes, then a
new visit will be reported if another
interaction occurs.
Sum of unique Google Analytics
tracking code and browser
cookies that visit a website at least
once during a measurement
period.
Ratio of single-page visits divided by
all visits to a website during a
measurement period
Single-page sessions have an
undefined session duration, since
there are no subsequent server hits
after the first one that would let
Analytics calculate the length of the
session. However, using a period of
inactivity for the exit, bounce
sessions have a duration of zero.
Session duration is the period of a
group of user interactions with a
website from the first and
subsequent interactions to a period
of inactivity. By default, a session
lasts until there are 30 minutes of
inactivity.
Session duration relies on a period
of inactivity to end the session, as
there is no server hit when the
visitor exits the website.
SimilarWeb Sum of times at least one page of a
website has been loaded into a
browser during a measurement
period Subsequent page views are
included in the same visit until the
user is inactive for more than 30
minutes. If a user becomes active
again after 30 minutes, that counts as
a new visit. A new session will also
start at midnight.
Sum of computing devices
visiting a website within a
geographical area and during a
measurement period.
Ratio of single page visits by all visits
for a website within a geographical
area and during a measurement
period.
Session duration is the period of is a
group of user interactions with a
website from the first and
subsequent interactions to a period
of inactivity. By default, a session
lasts until there are 30 minutes of
inactivity.
https://doi.org/10.1371/journal.pone.0268212.t002
Table 3. Host country of organization for 86 websites in study.
Country No. %
United States 43 50.0%
India 6 7.0%
Russian Federation 6 7.0%
Japan 4 4.7%
United Kingdom 4 4.7%
France 3 3.5%
Israel 3 3.5%
Spain 2 2.3%
One each (Belarus, Belgium, Canada, Chile, China, Cuba, Ecuador, Germany, Madagascar, Malaysia,
Nigeria, Taiwan, Turkey, Ukraine, United Arab Emirates)
15 17.4%
86 100.0%
https://doi.org/10.1371/journal.pone.0268212.t003
PLOS ONE Measuring user interactions with websites
PLOS ONE | https://doi.org/10.1371/journal.pone.0268212 May 27, 2022 10 / 27
underestimated by 643 million (19.4%) total visits. Using Google Analytics numbers as the
baseline for total visits, SimilarWeb overestimated 15 (17.4%) sites and underestimated 66
(76.7%) sites. The two platforms were nearly similar (~+/- 5%) for 5 (5.8%) sites.
Ranking the websites by total visits based on Google Analytics and SimilarWeb, we then
conduct a Pearson correlation coefficient test. There was a significant strong positive associa-
tion between the ranking of Google Analytics and SimilarWeb, rs(85) = .954, p< .001.
Graphically, we compare the reported total visits between Google Analytics and Similar-
Web in Fig 2, showing the correlational relationship. As shown in Fig 2, the number of total
visits between Google Analytics and SimilarWeb has a strong, positive, linear correlation.
This finding implies that, although the reported total visits values differ between the two
platforms, the trend for the set of websites is generally consistent. So, if one is interested in a
ranking (e.g., “Where does website X rank within this set of websites based on total visits?”),
then SimilarWeb values will generally align with those of Google Analytics for those websites.
However, if one is specifically interested in numbers (e.g., “What is the number of total visits
to each of N websites?), then the SimilarWeb total visit numbers will be ~20% below those
reported by Google Analytics, on average.
Table 4. Industry vertical of organization for 86 websites in study.
Website Category No. %
News and Media 36 41.9%
Computers Electronics and Technology 10 11.6%
Arts and Entertainment 9 10.5%
Science and Education 5 5.8%
Community and Society 4 4.7%
Finance 4 4.7%
Business and Consumer Services 2 2.3%
E-commerce and Shopping 2 2.3%
Gambling 2 2.3%
Travel and Tourism/ 2 2.3%
Vehicles 2 2.3%
Health 1 1.2%
Hobbies and Leisure 1 1.2%
Home and Garden 1 1.2%
Jobs and Career 1 1.2%
Law and Government 1 1.2%
Lifestyle/Beauty and Cosmetics 1 1.2%
Lifestyle/Fashion and Apparel 1 1.2%
Sports 1 1.2%
86 100.0%
https://doi.org/10.1371/journal.pone.0268212.t004
Table 5. Website type for the 86 websites in study.
Site Type No. %
Content 50 58.1%
Other 34 39.5%
Transactional 2 2.3%
86 100.0%
https://doi.org/10.1371/journal.pone.0268212.t005
PLOS ONE Measuring user interactions with websites
PLOS ONE | https://doi.org/10.1371/journal.pone.0268212 May 27, 2022 11 / 27
H2: Measurements of unique visitors differ
A paired t-test was conducted to compare the number of unique visitors reported by Google
Analytics and SimilarWeb. There was a significant difference in unique visitors for the Google
Analytics (M = 6.56, SD = 0.26 million) and the SimilarWeb (M = 6.31, SD = 0.25) conditions;
t(85) = 12.60, p< 0.01. These results indicate a difference in the number of unique visitors
between the two approaches. Specifically, our results show that the reported number of unique
visitors by SimilarWeb is statistically lower than the values reported by Google Analytics.
Therefore, H2 is fully supported: SimilarWeb measurement of unique visitors to websites
differ from those reported by Google Analytics.
The total number of unique visitors for all 86 websites was 834.7 million (max = 138.1 mil-
lion; min = 1,799; med = 4.3 million) reported by Google Analytics and 439.0 million
(max = 54.6 million; min = 2,361; med = 2.3 million) reported by SimilarWeb. Using the mean
aggregate unique visitors for all 86 websites, using Google Analytics as the baseline, SimilarWeb
underestimated by 395.6 million (38.7%) unique visitors. Using Google Analytics numbers as
the baseline, SimilarWeb overestimated 4 (4.7%) sites and underestimated 82 (95.3%) sites.
Ranking the websites by unique visitors based on Google Analytics and SimilarWeb, we
then conduct a Pearson correlation coefficient test. There was a significant strong positive
association between the ranking of Google Analytics and SimilarWeb, rs(85) = .967, p< .001.
Graphically, we compare the reported unique visitors between Google Analytics and Simi-
larWeb in Fig 3, showing the correlational relationship. As shown in Fig 3, the number of total
visits between Google Analytics and SimilarWeb has a strong, positive, linear correlation.
This finding indicates that, while the reported values for unique visitors differ between the
two platforms, the trend for the set of websites is mostly consistent. So, if one is interested in a
ranking (e.g., “Where does website X rank within this set of websites based on unique visi-
tors?”), then SimilarWeb values will generally align with those of Google Analytics for those
websites. However, if one is specifically interested in numbers (e.g., “What is the number of
unique visitors to each of N websites?), then the SimilarWeb unique visitor numbers will be
~40% below those reported by Google Analytics, on average.
Fig 2. Scatter plot of total visits reported by Google Analytics and SimilarWeb showing strong, positive, linear
correlation.
https://doi.org/10.1371/journal.pone.0268212.g002
PLOS ONE Measuring user interactions with websites
PLOS ONE | https://doi.org/10.1371/journal.pone.0268212 May 27, 2022 12 / 27
H3: Measurements of bounce rates differ
A paired t-test was conducted to compare bounce rates reported by Google Analytics and Simi-
larWeb. There was a significant difference in the bounce rates between the Google Analytics
(M = 0.58, SD = 0.03) and the SimilarWeb (M = 0.63, SD = 0.02) conditions; t(85) = -2,96,
p< 0.01. Specifically, our results showed that the reported bounce rate by SimilarWeb was sig-
nificantly higher than that reported by Google Analytics, fully supporting H3: SimilarWeb
measurement of bounce rates for websites differ from those reported by Google Analytics.
The average of bounce rate for all 86 websites was 56.2% (SS = 20.4%, max = 88.9%;
min = 20.4%; med = 59.2%) reported by Google Analytics and 63.0% (SS = 13.8%,
max = 86.0%; min = 28.8%; med = 65.3%) as reported SimilarWeb. Using Google Analytics as
the baseline, SimilarWeb analytics were 6.8% more than the average bounce rate. Additionally,
SimilarWeb over calculated 35 (40.7%) sites and under calculated 31 (36.0%) sites. The two
platforms were nearly similar (+/- 5) for 20 (23.3%) sites.
We then conducted a Pearson correlation coefficient test to rank the websites by bounce
rate based on Google Analytics and SimilarWeb. There was a significant positive association
between the ranking of Google Analytics and SimilarWeb, rs(85) = .461, p< .001.
Graphically, this is illustrated in Fig 4, where we compare bounce rates between Google
Analytics and SimilarWeb. As shown in Fig 4, the bounce rates between Google Analytics and
SimilarWeb have a moderate, positive, linear correlation.
This finding indicates that, although SimilarWeb and Google Analytics report similar
bounce rates for more than 20% of the sites, the difference between the values for the other
80% for the two platforms was high. We address the possible reasons for this high discrepancy
later in the discussion of results.
H4: Measurements of average session duration differ
A paired t-test was conducted to compare the average session duration reported by Google
Analytics and SimilarWeb. There was a significant difference in the average session duration
Fig 3. Scatter plot of unique visitors reported by Google Analytics and SimilarWeb showing strong, positive,
linear correlation.
https://doi.org/10.1371/journal.pone.0268212.g003
PLOS ONE Measuring user interactions with websites
PLOS ONE | https://doi.org/10.1371/journal.pone.0268212 May 27, 2022 13 / 27
between the Google Analytics (M = 2.15, SD = 0.05) and the SimilarWeb (M = 2.47, SD = 0.71)
conditions; t(85) = -8.59, p< 0.01. Specifically, our results showed that the reported average
session duration by SimilarWeb was significantly higher than that reported by Google Analyt-
ics, fully supporting H4: SimilarWeb measurement of average session duration for web-
sites differ from those reported by Google Analytics.
The average session duration for all 86 websites was 202.91 seconds (SS = 239.71,
max = 1439.51; min = 33.25; med = 119.63) reported by Google Analytics and 463.51 seconds
(SS = 640.99, max = 4498.08; min = 62.42; med = 267.13) as reported SimilarWeb. Using Goo-
gle Analytics as the baseline, SimilarWeb reported a 52.6% more total average session duration.
Additionally, SimilarWeb over reported 63 (73.3%) sites and under reported 9 (10.5%) sites,
relative to Google Analytics. The two platforms were nearly similar (~+/- 5) for 14 (16.3%)
sites.
Ranking the websites by average session duration based on Google Analytics and Similar-
Web, we then conduct a Pearson correlation. There was a significant positive association
between the ranking of Google Analytics and SimilarWeb, rs(85) = .536, p< .001, as shown in
Fig 5.
This finding indicates that, although SimilarWeb and Google Analytics report similar aver-
age sessions for about 16% of the sites, the difference between the values for the other 84% of
the sites for the two platforms was generally high. We address the possible reasons for this high
discrepancy later in the discussion of results.
Discussion
General discussion
Table 6 summarizes our findings for the 86 websites using average monthly total visits, unique
visitors, bounce rate, and average session duration during the 12-month analysis period.
As shown in Table 6, statistical testing of all four hypotheses is statistically significant, so all
four hypotheses are supported. The reported values for total visits, unique visitors, bounce
Fig 4. Scatter plot of bounce rate reported by Google Analytics and SimilarWeb showing moderate, positive,
linear correlation.
https://doi.org/10.1371/journal.pone.0268212.g004
PLOS ONE Measuring user interactions with websites
PLOS ONE | https://doi.org/10.1371/journal.pone.0268212 May 27, 2022 14 / 27
rates, and average session duration for Google Analytics and SimilarWeb differ significantly.
The website rankings by each service are significantly correlated, so it seems that these ranked
lists can be used for research on analytics, competitive analysis, and analytics calculations for a
set of websites, with the caveat highlighted in [18, 19]. These analyses compare the two ser-
vices’ precision (i.e., how close measured values are to each other).
However, the underlying question motivating our research remains this: How accurate are
the reported metrics from website analytics services (i.e., how close are the reported values to the
‘true’ values)? Regardless of the statistical testing results, this motivational question is more
challenging to address. In reality, there is one ‘true’ number of visits, visitors, bounces, and
average session duration. However, is it realistic to expect any web analytics service to match
reality perfectly? Moreover, what is the reality in terms of web analytics? In our perspective, it
is a misconception to view web analytics data collection as “counting.” In most cases, web ana-
lytics is not counting; instead, it is “measuring.” It is well known that there will be an error rate
Fig 5. Scatter plot of average session duration reported by Google Analytics and SimilarWeb showing moderate,
positive, linear correlation.
https://doi.org/10.1371/journal.pone.0268212.g005
Table 6. Summary of results comparing Google Analytics and SimilarWeb for total visits, unique visitors, bounce rate, and average session duration. Difference
uses Google Analytics as the Baseline. Results based on Paired t-Test for Hypotheses Supported.
Metric / Service Google Analytics SimilarWeb Difference Hypotheses
Total Visits 1,703,584,207 1,060,137,189 19.4% Fully Supported–The reported values differ
Unique Visitors 834,656,530 439,016,436 38.7% Fully Supported–The reported values differ
Bounce Rate 56.2% 63.0% 6.8% Fully Supported–The reported values differ
Average Session Duration 202.91 463.51 56.2% Fully Supported–The reported values differ
Number of Sites (Relative to Google Analytics Values) Where SimilarWeb Numbers Were:
Higher Lower Similar (~+/- 5%)
Total Visits 15 (17.4%) 66 (76.7%) 5 (5.8%) SimilarWeb values will generally be lower than Google Analytics
Unique Visitors 4 (4.7%) 82 (95.3%) 0 (0.0%) SimilarWeb values will generally be lower than Google Analytics
Bounce Rate 35 (40.7%) 31 (36.0%) 20 (23.3%) SimilarWeb values will generally be higher than Google Analytics
Average Session Duration 63 (73.3%) 9 (10.5%) 14 (16.3%) SimilarWeb values will generally be higher than Google Analytics
https://doi.org/10.1371/journal.pone.0268212.t006
PLOS ONE Measuring user interactions with websites
PLOS ONE | https://doi.org/10.1371/journal.pone.0268212 May 27, 2022 15 / 27
(+/- n%) for nearly any measure [117]. No measure or measurement tool is perfect, and web
data can be particularly messy.
Although one might lean toward considering metrics reported by Google Analytics as the
‘gold standard’ for website analytics (and justifiably so in many cases), it is also known within
the industry that Google Analytics has tracking issues in some cases. Also, a reportedly high
percentage of Google Analytics accounts are incorrectly set up [118–121], perhaps skewing the
measuring in some cases. There are also contexts where other analytics methods might be
more appropriate. Google Analytics relies on one data collection approach: basically, a cookie
and tagging technique. There are certainly cases (e.g., cleared cookies, incognito browsing)
when this method is inaccurate (e.g., unique visitors). Furthermore, Google Analytics might
have different settings in terms of filtering, such as housekeeping visits from organizational
employees that would slant the results. Therefore, these concerns result in issues with Google
Analytics being seen as the ‘gold standard.’
To investigate our motivation research question regarding the accuracy of Google Analytics
and SimilarWeb as analytics services, we conduct a deductive analysis using a likelihood of
error [122]. We analyze what makes theoretical sense for which web analytics approach, Goo-
gle Analytics or SimilarWeb, would result in the most accurate measurement for each of our
metrics. We discuss our analysis of each metric below.
Bounce rate (engagement). A high bounce rate is undesirable for many sites. The bounce
rate means that someone comes to a site and leaves without taking relevant action. For this
metric, both Google Analytics [123] and SimilarWeb are conceptually incorrect due to the
practical issues of measuring a bounce visit [124]. For a meaningful session measurement,
there must be an entry point (where the person came to the site) and an exit point (where the
person left the site). If there is no endpoint to the session, both Google Analytics and Similar-
Web count it as a single page visit and a bounce because there is no exit interaction.
There are many situations where relevant action is taken on a site, but there is no exit point
[125]. For example, there can be an e-commerce site where a potential consumer arrives on a
product page, reads the content, and takes no other action at that time. Another case is a news-
paper site where an audience member comes to the site, scans the headlines, reads the article
snippets, but takes no other action, such as clicking [126]. In each of these cases, the visit could
last several minutes or longer. However, since there is no exit page (i.e., no second page), Goo-
gle Analytics and SimilarWeb would count these example visits as bounces.
So, we can reasonably assume both Google Analytics and SimilarWeb are overcounting
bounces, conceptually. This may be why the values vary substantially between the two services.
However, since bounce is a site-centric specific measure, we would expect Google Analytics to
be more precise (if not more accurate) than SimilarWeb when measuring bounce rate on a sin-
gle given site. However, SimilarWeb’s panel data may help correct this somewhat for a set of
sites, which Google Analytics does not measure. So, if one needs to examine the bounce rate of
several websites, Google Analytics cannot be used since website owners usually do not make
their web analytics data available to the public.
In terms of mechanical metrics, one would expect Google Analytics to be better for an indi-
vidual site. SimilarWeb might be expected to give reasonable bounce rate numbers for some
sites due to their user-centric panel data, and bounce rates are generally high, especially for
highly trafficked sites. This reasonableness in results from both Google Analytics and Similar-
Web is borne out in our statistical analysis above, where the two services were more in agree-
ment for the larger traffic sites for bounce rates (see Fig 4).
Average session duration (duration). Again, for this metric, both Google Analytics [123]
and SimilarWeb are conceptually incorrect due to the practical issues of measuring the end of
a session. Similar to the bounce rate, there is no exit point (i.e., where the person left the site).
PLOS ONE Measuring user interactions with websites
PLOS ONE | https://doi.org/10.1371/journal.pone.0268212 May 27, 2022 16 / 27
As there is no endpoint, both Google Analytics and SimilarWeb rely on a temporal timeout
measured from the time of the last interaction. So, this most likely under measures the dura-
tion of many sessions.
Again, since average session duration is a site-centric specific measure, Google Analytics
would be expected to be at least more precise (if not more accurate) than SimilarWeb when
measuring average session duration on a single given site. Again, SimilarWeb’s panel data may
somewhat help correct this for a set of sites for which Google Analytics data is unavailable. So,
similar to bounce rate, if one needs to examine the average session duration of several websites,
Google Analytics cannot be used, as this data is usually not public. In the end, conceptually,
both Google Analytics and SimilarWeb are most likely under measuring average session dura-
tion. In terms of practical implementation, one would expect Google Analytics to be better for
an individual site. SimilarWeb might be expected to give reasonable numbers for some sites
due to their user-centric panel data.
Total visits (frequency). This seems like a straightforward site-centric metric for which
Google Analytics should excel. Although there is room for some noise in the visits, such as
housekeeping visits (i.e., visits from internal company personnel for site maintenance), bot-
generated visits [127], purchased traffic, or hacking attacks that might not conceptually meet
the definition of a visit, it is difficult to imagine how an analytics service could be better than a
site-centric service in this regard. Using the site and network-centric data collection data
employed by website analytics services like SimilarWeb would not mitigate some of the noise
mentioned above; however, the user-centric panel data might compensate for some of the
noise issues for at least a high traffic website and for bot traffic. However, in general, one
would expect Google Analytics to be more accurate in measuring visits than SimilarWeb.
However, Google Analytics data is generally unavailable for multiple websites, so relying on
Google Analytics is not practical for these situations. For these cases, one would need to
employ an analytics service, such as SimilarWeb. Based on our analysis above, values for total
visits from SimilarWeb would be less than Google Analytics measurements by ~20% on
average.
Unique visitors (reach). Finally, we consider unique visitors. In this case, perhaps surpris-
ingly, one would expect the greater likelihood of error to be with the site-centric measure-
ments, resulting in SimilarWeb measures being more accurate.
Site-centric services, such as Google Analytics, typically rely on a combination of cookies
and tags to measure unique visitors. This approach would generally result in an overcount of
unique visitors by the service. For example, the expected life cycle of a computer is three to five
years [128, 129], meaning a person changing computers would be registered as a new visitor.
The market share of browsers has changed considerably over the years [130, 131], meaning
when someone has changed browsers, he/she would be registered as a new visitor. Studies
show that 40% of Internet users clear cookies daily, weekly, or monthly [132, 133], and about
3.7% of users disable all cookies [134, 135]. These actions would trigger a unique visitor count
when visiting a website. Some studies point to a much higher rate, with more than 30% of
users deleting cookies in a given month [132]. Many people also use the incognito mode on
their browsers [136, 137], triggering a new visitor count in Google Analytics [138, 139]. Also,
many people have multiple devices (e.g., personal computer, work computer, smartphone, tab-
let), with about 50% of Americans, for example, using four Internet-enabled devices [140,
141], so each device would be counted as a unique visitor even if it is the same person is using
the multiple devices.
For these reasons, the unique visitor number measured using the cookie approach would
likely lead to an overcount using site-centric metrics. Howmuch of an overcount? Based on the
issues just outlined, it seems that, for Google Analytics, a 20% overestimate in monthly unique
PLOS ONE Measuring user interactions with websites
PLOS ONE | https://doi.org/10.1371/journal.pone.0268212 May 27, 2022 17 / 27
visitors to 30% overestimate for more extended periods seems reasonable. However, more pre-
cise measures require an in-depth study and are a task for future research.
For unique visitors, it seems that panel data, such as those that Similar Web and other net-
work-centric services use, might be more accurate. However, this might only hold for larger
websites. It is not clear that panel data would be accurate for lower-traffic websites as there is
not enough traffic to these sites to generate reasonable statistical analysis. Generally, for unique
visitors, it seems that Google Analytics would most likely overestimate the number of unique
visitors to the website. SimilarWeb might be more accurate for the larger traffic websites due
to its user panel data approach but have questionable accuracy (either over- or underestimat-
ing) for the smaller traffic websites. Again, this conclusion is borne out by our analysis above,
where the difference between Google Analytics and SimilarWeb increased for the smaller web-
sites (see Fig 3).
Theoretical implications
We highlight three theoretical implications of this research, which are:
• Triangulation of Data, Methods, and Services: There seems, at present, to be no single data
collection approach (user, site, or network-centric) or web analytics service (including Goo-
gle Analytics or SimilarWeb) that would be effective for all metrics, contexts, or business
needs. Therefore, a triangulation of services, depending on the data, method of analysis, or
need, seems to be the most appropriate approach. It appears reasonable that user-centric
approaches can be leveraged for in-depth investigation of user online behaviors, albeit usu-
ally with a sample. Site-centric approaches can be leveraged for the investigation of users’
onsite behaviors. Network-centric approaches can be leveraged for in-depth investigation of
user intersite behaviors (i.e., navigation between sites).
• Discrepancies with Implementation: Regarding precision, we have established differences
between the two services, and we know the general methodologies and metrics calculations.
However, the nuances of implementation have not been independently audited as of the
date of this study. So, in practice, we cannot say definitely which is the best implementation
for a given metric. Again, this points to the need for triangulation of methods. It also high-
lights the lack of a gold standard for evaluating website analytics services. Regardless of any
nuances in implementation, the values between the two services are correlated, and, as dis-
cussed above, we can infer the preferred approach using deductive analysis.
• Discrepancies with Reality: Precision does not mean accuracy for either Google Analytics
or SimilarWeb. We have already outlined potential issues with all four of the metrics exam-
ined (i.e., total visits, unique visitors, bounce rates, average session duration). The applica-
tion mechanics are not aligned with conceptual definitions of what these metrics supposedly
measure. This situation calls for both continued research into improved measures and a real-
ization that the reported values (from both Google Analytics and SimilarWeb) are not counts
per se and should not be viewed necessarily as ‘truth.’ Rather, the values are reported mea-
sures with some error rates (+/-).
Practical implications
We highlight three practical implications of the findings, which are:
• Use of Google Analytics and SimilarWeb: Findings of our research show that, in general,
SimilarWeb results for total visits and number of unique visitors will generally be lower than
those reported by Google Analytics, and the correlation between the two platforms is high
PLOS ONE Measuring user interactions with websites
PLOS ONE | https://doi.org/10.1371/journal.pone.0268212 May 27, 2022 18 / 27
for these two metrics. So, if one is interested in ranking a set of websites for which one does
not have the Google Analytics data, the SimilarWeb metrics are a workable proxy. If one is
interested in the actual Google Analytics traffic for a set of websites, one can use the Similar-
Web results and increase by about 20% for total visits and about 40% for unique visitors, on
average. As a caveat, the Google Analytics unique visitor’s numbers are probably an over-
count, and the SimilarWeb values may be more in line with reality. As an easier ‘rule of
thumb’, we suggest using a 20% adjustment (i.e., increase SimilarWeb numbers) for both
metrics based on the analysis findings above. The realization that these services can be com-
plementary can improve decision-making that relies on KPIs and metrics from website ana-
lytics data.
• Verification of Analytics for a Single Website: In general, Google Analytics is a site-centric
web analytics platform, so it would be a reasonable service to use for a single website that
one owns and has access. However, comparing analytics values from Google Analytics to
those of SimilarWeb (or other website analytics services) may be worthwhile, as these will be
the values that outsiders see concerning the website.
• Estimating Google Analytics Metrics for Multiple Websites: As shown above, the differ-
ences between Google Analytics and SimilarWeb metrics for total visits and unique visitors
are systematic (i.e., the differences stay relatively constant), notably for visits and unique visi-
tors. This means, if you have Google Analytics values for one site, you can adjust and use a
similar difference for the other websites to get reasonable analytics numbers to those from
Google Analytics. This technique is valuable in competitive analysis situations where you
compare multiple sites against a known website and want the Google Analytics values for all
sites. However, SimilarWeb generally provides conservative analytics metrics compared to
Google Analytics, meaning that, if solely relying on this single service, analytics measures
may be lower, especially for onsite interactions. So, decisions using these analytics metrics
need to include this as a factor.
Limitations, future research, and strengths
Limitations and future research. The first limitation concerns data quality. In the
absence of ground truth, we primarily measure precision and not the accuracy of the two web
analytics services. As noted, there are inconsistencies between the two platforms. So, the ana-
lytics data that decision-makers may perceive as accurate, objective, and correct may not have
these qualities due to the several potential sources for errors outlined above. Web analytics ser-
vices should undertake future research to provide metric values with confidence intervals to
depict them as ranges rather than exact values. Another limitation is that the source codes and
specific implementations for either of these platforms are not available, so the nuances of the
implementations cannot be verified. Although it is apparent from results and from company
materials that both platforms use state-of-the-art algorithmic approaches, future research
could focus on using open-source analytics platforms, such as Matomo [142], to tease apart
some of these metric implementations. An additional limitation is that a large percentage of
the sites used in this research are content creation sites based in the U.S.A., which might skew
user behavior. Other future research involves replication studies with different sets of websites,
other website analytics services, other metrics, and analysis of specific website segments based
on type, size, industry vertical, or country (i.e., China being a critical region of interest).
Strengths. There are several strengths of this research. First, we use two popular web ana-
lytics services. Second, we employ 86 websites with various attributes, ensuring a robust sam-
ple size. Third, we collect data over an extended period of 12 months to mitigate for short
PLOS ONE Measuring user interactions with websites
PLOS ONE | https://doi.org/10.1371/journal.pone.0268212 May 27, 2022 19 / 27
periods of fluctuation with the website analytics measures. Fourth, we report and statistically
evaluate four core web analytics metrics–total visits, unique visitors, bounce rates, and average
session duration. Fifth, we discuss and offer theoretical and practical implications of our
research. To our knowledge, this is one of the first and one of the most extensive academic
examinations and analyses of these popular web analytics services.
Conclusion
For this research, we compared four analytics metrics from Google Analytics to those from
SimilarWeb based on 12 months of data for 86 diverse websites. Findings show statistically sig-
nificant differences between the two services for total visits, unique visitors, bounce rates, and
average session duration. Compared to Google Analytics, SimilarWeb values were ~20% lower
for total visits, ~40% lower for unique visitors, ~25% higher for bounce rate, and ~50% higher
for average session duration, on average. The rankings of all four metrics are significantly cor-
related between Google Analytics and SimilarWeb, and the measurement differences are sys-
tematic between the two analytics services. The implications are that SimilarWeb provides
conservative analytics results relative to Google Analytics, and these web analytics tools can be
complementarily utilized in various contexts, especially when having data for one website and
needing analytics data for other websites.
Supporting information
S1 File.
(DOCX)
Author Contributions
Conceptualization: Bernard J. Jansen, Joni Salminen.
Formal analysis: Bernard J. Jansen.
Investigation: Bernard J. Jansen.
Methodology: Bernard J. Jansen.
Writing – original draft: Bernard J. Jansen.
Writing – review & editing: Bernard J. Jansen, Soon-gyo Jung, Joni Salminen.
References
1. Jansen BJ. Understanding User-Web Interactions via Web Analytics. Synthesis Lectures on Informa-
tion Concepts, Retrieval, and Services. 2009 Jan 1; 1(1):1–102.
2. Saura JR. Using Data Sciences in Digital Marketing: Framework, methods, and performance metrics.
Journal of Innovation & Knowledge. 2021; 6(2):92–102.
3. Jiang J, He D, Allan J. Searching, browsing, and clicking in a search session: changes in user behavior
by task and over time. In: Proceedings of the 37th international ACM SIGIR conference on Research
& development in information retrieval. New York, NY, USA: Association for Computing Machinery;
2014. p. 607–16. (SIGIR ‘14).
4. Ka¨mpf M, Tessenow E, Kenett DY, Kantelhardt JW. The Detection of Emerging Trends Using Wikipe-
dia Traffic Data and Context Networks. PLOS ONE. 2015 Dec 31; 10(12):e0141892. https://doi.org/
10.1371/journal.pone.0141892 PMID: 26720074
5. Choo C, Detlor B, Turnbull D. Information Seeking on the Web: An Integrated Model of Browsing and
Searching. First Monday. 2000 Mar 27; 5.
6. Figueredo de Santana V, Ferreira Silva FE. User Test Logger: An Open Source Browser Plugin for
Logging and Reporting Local User Studies. In: Antona M, Stephanidis C, editors. Universal Access in
PLOS ONE Measuring user interactions with websites
PLOS ONE | https://doi.org/10.1371/journal.pone.0268212 May 27, 2022 20 / 27
Human-Computer Interaction Theory, Methods and Tools. Cham: Springer International Publishing;
2019. p. 229–43. (Lecture Notes in Computer Science).
7. Jansen BJ, McNeese MD. Evaluating the effectiveness of and patterns of interactions with automated
searching assistance. Journal of the American Society for Information Science and Technology. 2005;
56(14):1480–503.
8. Miroglio B, Zeber D, Kaye J, Weiss R. The Effect of Ad Blocking on User Engagement with the Web.
In: Proceedings of the 2018 World Wide Web Conference. Republic and Canton of Geneva, CHE:
International World Wide Web Conferences Steering Committee; 2018. p. 813–21. (WWW ‘18).
9. Ahmed H, Tahseen D, Haider W, Asad M, Nand S, Kamran S. Establishing Standard Rules for Choos-
ing Best KPIs for an E-Commerce Business based on Google Analytics and Machine Learning Tech-
nique. International Journal of Advanced Computer Science and Applications. 2017 Jan 1; 8.
10. Gunter U, O¨ nder I. Forecasting city arrivals with Google Analytics. Annals of Tourism Research. 2016
Nov 1; 61:199–212.
11. He D, Go¨ker A, Harper DJ. Combining evidence for automatic Web session identification. Information
Processing & Management. 2002 Sep 1; 38(5):727–42.
12. Jiang T, Chi Y, Gao H. A clickstream data analysis of Chinese academic library OPAC users’ informa-
tion behavior. Library & Information Science Research. 2017 Jul 1; 39(3):213–23.
13. Jiang T, Yang J, Yu C, Sang Y. A Clickstream Data Analysis of the Differences between Visiting
Behaviors of Desktop and Mobile Users. Data and Information Management. 2018 Dec 31; 2(3):130–
40.
14. Ortiz-Cordova A, Jansen BJ. Classifying Web Search Queries in Order to Identify High Revenue Gen-
erating Customers. Journal of the American Society for Information Sciences and Technology. 2012;
63(7):1426–41.
15. Vecchione A, Brown D, Allen E, Baschnagel A. Tracking User Behavior with Google Analytics Events
on an Academic Library Web Site. Journal of Web Librarianship. 2016 Jul 2; 10(3):161–75.
16. Wang P, Berry MW, Yang Y. Mining longitudinal web queries: trends and patterns. Journal of the
American Society for Information Science and Technology. 2003; 54(8):743–58.
17. Midha V. The Glitch in On-line Advertising: A Study of Click Fraud in Pay-Per-Click Advertising Pro-
grams. International Journal of Electronic Commerce. 2008 Dec 1; 13(2):91–112.
18. Pochat VL, Van Goethem T, Tajalizadehkhoob S, Korczyński M, Joosen W. Tranco: A Research-Ori-
ented Top Sites Ranking Hardened Against Manipulation. Proceedings 2019 Network and Distributed
System Security Symposium. 2019;
19. Scheitle Q, Hohlfeld O, Gamba J, Jelten J, Zimmermann T, Strowes SD, et al. A Long Way to the Top:
Significance, Structure, and Stability of Internet Top Lists. Proceedings of the Internet Measurement
Conference 2018. 2018;478–93.
20. Martı´n-Martı´n A, Orduna-Malea E, Thelwall M, Delgado Lo´pez-Co´zar E. Google Scholar, Web of Sci-
ence, and Scopus: A systematic comparison of citations in 252 subject categories. Journal of Infor-
metrics. 2018 Nov 1; 12(4):1160–77.
21. Thelwall M. Interpreting social science link analysis research: A theoretical framework. Journal of the
American Society for Information Science and Technology. 2006; 57(1):60–8.
22. Similarweb. SimilarWeb Data Methodology [Internet]. SimilarWeb Data Methodology. 2022 [cited
2022 Feb 2]. Available from: http://support.similarweb.com/hc/en-us/articles/360001631538
23. Hang H, Bashir A, Faloutsos M, Faloutsos C, Dumitras T. “Infect-me-not”: A user-centric and site-cen-
tric study of web-based malware. In: 2016 IFIP Networking Conference (IFIP Networking) and Work-
shops. 2016. p. 234–42.
24. Prantl D, Prantl M. Website traffic measurement and rankings: competitive intelligence tools examina-
tion. International Journal of Web Information Systems. 2018; 14(4):423–37.
25. Shukla A, Gullapuram SS, Katti H, Yadati K, Kankanhalli M, Subramanian R. Evaluating content-cen-
tric vs. user-centric ad affect recognition. In: Proceedings of the 19th ACM International Conference
on Multimodal Interaction. New York, NY, USA: Association for Computing Machinery; 2017. p. 402–
10. (ICMI ‘17).
26. Zheng Z (Eric), Fader P, Padmanabhan B. From Business Intelligence to Competitive Intelligence:
Inferring Competitive Measures Using Augmented Site-Centric Data. Information Systems Research.
2011 Nov 3; 23(3-part-1):698–720.
27. Croll A, Power S. Complete Web Monitoring: Watching your visitors, performance, communities, and
competitors. O’Reilly Media, Inc.; 2009. 666 p.
PLOS ONE Measuring user interactions with websites
PLOS ONE | https://doi.org/10.1371/journal.pone.0268212 May 27, 2022 21 / 27
28. Nepusz T, Petro´czi A, Naughton DP. Network Analytical Tool for Monitoring Global Food Safety High-
lights China. PLOS ONE. 2009 Aug 18; 4(8):e6680. https://doi.org/10.1371/journal.pone.0006680
PMID: 19688088
29. Shawki Amin. 6 Free Analytics Tools to Help You Understand Your Competitor’s Web Traffic [Internet].
InfoTrust. 2013 [cited 2022 Feb 2]. Available from: https://infotrust.com/articles/6-free-analytics-tools-
to-help-you-understand-your-competitor-s-web-traffic/
30. Joint Laura. Metrics we measure • PR Resolution—by CoverageBook [Internet]. PR Resolution—by
CoverageBook. 2016 [cited 2022 Feb 2]. Available from: https://resolution.coveragebook.com/
metrics-we-measure/
31. Macanas Mark. SimilarWeb vs Google Analytics Traffic Data Mismatch, Explained [Internet]. TechPi-
nas: Philippines’ Technology News, Tips and Reviews Blog. [cited 2022 Feb 2]. Available from: https://
www.techpinas.com/2018/06/SimilarWeb-vs-Google-Analytics.html
32. Novak K. SimilarWeb vs Alexa: Which Traffic Estimator is More Precise? [Internet]. Growtraffic Blog.
2019 [cited 2022 Feb 2]. Available from: https://growtraffic.com/blog/2019/03/similarweb-alexa-which-
precise
33. Finley Olly. SEMrush vs SimilarWeb: What is the Best Tool for Media Buyers? [Internet]. Blog lemon-
ads. 2020 [cited 2022 Feb 2]. Available from: https://www.lemonads.com/blog/semrush-vs-
similarweb-what-is-the-best-tool-for-media-buyers/
34. Hinkis Roy. Traffic and Engagement Metrics and Their Correlation to Google Rankings [Internet]. Moz.
[cited 2022 Feb 2]. Available from: https://moz.com/blog/traffic-engagement-metrics-their-correlation-
to-google-rankings
35. Tyler Horvath. 8 Most Accurate Website Traffic Estimators [Internet]. Ninja Reports. 2020 [cited 2022
Feb 2]. Available from: https://www.ninjareports.com/website-traffic-estimators/
36. Barzilay O. How SimilarWeb Helps Investors Make Decisions About Their Portfolio [Internet]. Forbes.
2017 [cited 2022 Feb 2]. Available from: https://www.forbes.com/sites/omribarzilay/2017/11/09/meet-
similarweb-one-of-wall-streets-secret-weapons/
37. Bakaev M, Khvorostov V, Heil S, Gaedke M. Web Intelligence Linked Open Data for Website Design
Reuse. In: Cabot J, De Virgilio R, Torlone R, editors. Web Engineering. Cham: Springer International
Publishing; 2017. p. 370–7. (Lecture Notes in Computer Science).
38. Ng YMM, Taneja H. Mapping User-Centric Internet Geographies: How Similar are Countries in Their
Web Use Patterns? Journal of Communication. 2019 Oct 1; 69(5):467–89.
39. SimilarWeb. Our Data | Similarweb [Internet]. Similarweb. 2022 [cited 2022 Jan 30]. Available from:
https://www.similarweb.com/corp/ourdata/
40. West BT, Sakshaug JW, Aurelien GAS. How Big of a Problem is Analytic Error in Secondary Analyses
of Survey Data? PLOS ONE. 2016 Jun 29; 11(6):e0158120. https://doi.org/10.1371/journal.pone.
0158120 PMID: 27355817
41. Zhang M. Understanding the relationships between interest in online math games and academic per-
formance. Journal of Computer Assisted Learning. 2015; 31(3):254–67.
42. Ailawadi KL, Farris PW. Managing Multi- and Omni-Channel Distribution: Metrics and Research Direc-
tions. Journal of Retailing. 2017 Mar 1; 93(1):120–35.
43. Movsisyan SA. Social media marketing strategy of Yerevan brandy company. Annals of Agrarian Sci-
ence. 2016 Sep 1; 14(3):243–8.
44. Leitner P, Grechenig T. Scalable Social Software Services: Towards a Shopping Community Model
Based on Analyses of Established Web Service Components and Functions. In: 2009 42nd Hawaii
International Conference on System Sciences. 2009. p. 1–10.
45. Kagan S, Bekkerman R. Predicting Purchase Behavior of Website Audiences. International Journal of
Electronic Commerce. 2018 Oct 2; 22(4):510–39.
46. Hazrati N, Ricci F. Recommender systems effect on the evolution of users’ choices distribution. Infor-
mation Processing & Management. 2022 Jan 1; 59(1):102766.
47. Karpf D. Social Science Research Methods in Internet Time. Information, Communication & Society.
2012 Jun 1; 15(5):639–61.
48. Diaz F, Gamon M, Hofman JM, Kıcıman E, Rothschild D. Online and Social Media Data As an Imper-
fect Continuous Panel Survey. PLOS ONE. 2016 Jan 5; 11(1):e0145406. https://doi.org/10.1371/
journal.pone.0145406 PMID: 26730933
49. The´ro H, Vincent EM. Investigating Facebook’s interventions against accounts that repeatedly share
misinformation. Information Processing & Management. 2022 Mar 1; 59(2):102804.
50. Albarran AB. The Social Media Industries. Routledge; 2013. 274 p.
PLOS ONE Measuring user interactions with websites
PLOS ONE | https://doi.org/10.1371/journal.pone.0268212 May 27, 2022 22 / 27
51. Bergh BGV, Lee M, Quilliam ET, Hove T. The multidimensional nature and brand impact of user-gen-
erated ad parodies in social media. International Journal of Advertising. 2011 Jan 1; 30(1):103–31.
52. Blombach A, Dykes N, Evert S, Heinrich P, Kabashi B, Proisl T. A New German Reddit Corpus. In:
Proceedings of the 15th Conference on Natural Language Processing (KONVENS 2019). 2019.
p. 278–9.
53. Blombach A, Dykes N, Heinrich P, Kabashi B, Proisl T. A Corpus of German Reddit Exchanges (GeR-
edE). In: Proceedings of the 12th Language Resources and Evaluation Conference. Marseille,
France: European Language Resources Association; 2020. p. 6310–6.
54. Nakayama M, Wan Y. A quick bite and instant gratification: A simulated Yelp experiment on consumer
review information foraging behavior. Information Processing & Management. 2021 Jan 1; 58
(1):102391.
55. Lund A, Zukerfeld M. Profiting from Open Audiovisual Content. In: Lund A, Zukerfeld M, editors. Cor-
porate Capitalism’s Use of Openness: Profit for Free? Cham: Springer International Publishing;
2020. p. 199–239. (Dynamics of Virtual Work).
56. Miller CC, Fox KA. 834: Americans view widely varied blog advice about home birth. American Journal
of Obstetrics and Gynecology. 2017 Jan; 216(1):S478–9.
57. Zhang M. Who are interested in online science simulations? Tracking a trend of digital divide in Internet
use. Computers & Education. 2014 Jul 1; 76:205–14.
58. Lasuin CA, Omar A, Ramayah T. Social media and brand engagement in the age of the customer. In
Kuching, Sarawak, MALAYSIA; 2015. p. 138–44.
59. Rosenblatt M, Curran T, Treiber J. Building Brands through Social Listening. In 2018. p. 71–4.
60. Tangmanee C. Comparisons of Website Visit Behavior between Purchase Outcomes and Product
Categories. International Journal of Research in Business and Social Science (2147–4478). 2017 Jul
19; 6(4):1–10.
61. Kang S. Factors influencing intention of mobile application use. International Journal of Mobile Com-
munications. 2014 Jan 1; 12(4):360–79.
62. Bansal H, Kohli S. Trust evaluation of websites: a comprehensive study. International Journal of
Advanced Intelligence Paradigms. 2019 Jan 1; 13(1–2):101–12.
63. Gunawan AB. Socialization of Terms of Use and Privacy Policy on Indonesian e-commerce Websites.
Journal of Social Science. 2020 Jul 26; 1(3):41–5.
64. Singal H, Kohli S. Trust necessitated through metrics: estimating the trustworthiness of websites. Pro-
cedia Computer Science. 2016; 85:133–40.
65. Singal H, Kohli S. Mitigating Information Trust: Taking the Edge off Health Websites. International
Journal of Technoethics. 2016 Jan 1; 7(1):16–33.
66. Weissbacher M. These Browser Extensions Spy on 8 Million Users Extended [Internet]. 2016 [cited
2022 Feb 2]. Available from: /paper/These-Browser-Extensions-Spy-on-8-Million-Users-Weissba-
cher/3fe57d1556158da7fe373fb577ac5cbbc3f1e84b
67. Chakrabortty K, Jose E. Relationship Analysis between Website Traffic, Domain Age and Google
Indexed Pages of E-commerce Websites. IIM Kozhikode Society & Management Review. 2018 Jul 1;
7(2):171–7.
68. Kro´l K, Halva J. Measuring Efficiency of Websites of Agrotouristic Farms from Poland and Slovakia.
Economic and Regional Studies / Studia Ekonomiczne i Regionalne. 2017 Jun 1; 10(2):50–9.
69. Vyas C. Evaluating state tourism websites using Search Engine Optimization tools. Tourism Manage-
ment. 2019 Aug 1; 73:64–70.
70. Akcan H, Suel T, Bro¨nnimann H. Geographic web usage estimation by monitoring DNS caches. In:
Proceedings of the first international workshop on Location and the web. New York, NY, USA: Associ-
ation for Computing Machinery; 2008. p. 85–92. (LOCWEB ‘08).
71. Bates S, Bowers J, Greenstein S, Weinstock J, Xu Y, Zittrain J. Evidence of Decreasing Internet
Entropy: The Lack of Redundancy in DNS Resolution by Major Websites and Services [Internet].
National Bureau of Economic Research; 2018 Feb [cited 2022 Feb 2]. (Working Paper Series). Report
No.: 24317. Available from: http://www.nber.org/papers/w24317
72. de Carlos P, Arau´jo N, Fraiz JA. The new intermediaries of tourist distribution: Analysis of online
accommodation booking sites. The International Journal of Management Science and Information
Technology. 2016;(19):39–58.
73. Das DB, Sahoo JS. Social Networking Sites–A Critical Analysis of Its Impact on Personal and Social
Life. International Journal of Business and Social Science. 2(14):222–8.
74. Marine-Roig E. A Webometric Analysis of Travel Blogs and Review Hosting: The Case of Catalonia.
Journal of Travel & Tourism Marketing. 2014 Apr 3; 31(3):381–96.
PLOS ONE Measuring user interactions with websites
PLOS ONE | https://doi.org/10.1371/journal.pone.0268212 May 27, 2022 23 / 27
75. Smoliarova AS, Gromova TM. News Consumption Among Russian-Speaking Immigrants in Israel
from 2006 to 2018. In: Alexandrov DA, Boukhanovsky AV, Chugunov AV, Kabanov Y, Koltsova O,
Musabirov I, editors. Digital Transformation and Global Society. Cham: Springer International Pub-
lishing; 2019. p. 554–64. (Communications in Computer and Information Science).
76. Suksida T, Santiworarak L. A study of website content in webometrics ranking of world university by
using similar web tool. In: 2017 IEEE 2nd International Conference on Signal and Image Processing
(ICSIP). 2017. p. 480–3.
77. Social Network Sites and Its Popularity. International Journal of Research and Reviews in Computer
Science. 2011; 2(2):522–6.
78. Lo B, Sedhain RS. How Reliable Are Website Rankings? Implications for E-Business Advertising and
Internet Search. 2006;(2):233–8.
79. Vaughan L, Yang R. Web traffic and organization performance measures: Relationships and data
sources examined. Journal of Informetrics. 2013 Jul 1; 7(3):699–711.
80. Nielsen. NetMonitor [Internet]. Nielsen Admosphere. 2015 [cited 2022 Feb 2]. Available from: https://
www.nielsen-admosphere.eu/products-and-services/internet-measurement/netmonitor/
81. Olivo Frank. Is SEMRush Accurate? A Comparison with My Site’s Analytics [Internet]. Sagapixel.
[cited 2022 Feb 2]. Available from: https://sagapixel.com/seo/is-semrush-accurate/
82. Mulder S. How to Compare Your Site Metrics to Your Local Competitors [Internet]. 2014 [cited 2022
Feb 2]. Available from: http://digitalservices.npr.org/post/how-compare-your-site-metrics-your-local-
competitors
83. Diachuk Olha, Loba Pavel, Mirgorodskaya Olga. Comparing accuracy: SEMrush vs SimilarWeb [Inter-
net]. owox. 2020 [cited 2022 Feb 2]. Available from: https://www.owox.com/blog/articles/semrush-vs-
similarweb/
84. Pupec Ioana. We analyzed 1787 eCommerce websites with SimilarWeb and Google Analytics and
that’s what we learned—Omniconvert Blog [Internet]. ECOMMERCE GROWTH Blog. 2017 [cited
2022 Feb 2]. Available from: https://www.omniconvert.com/blog/we-analyzed-1787-ecommerce-
websites-similarweb-google-analytics-thats-we-learned.html
85. Husain Osman. Forget the garage–this multi-million dollar company started in a jewelry store [Inter-
net]. Tech in Asia. 2015 [cited 2022 Feb 2]. Available from: https://www.techinasia.com/the-story-of-
similarweb
86. The SaaS Report. SimilarWeb | The Software Report [Internet]. n.d. [cited 2022 Feb 2]. Available
from: https://www.thesoftwarereport.com/top-companies/similarweb/
87. Hardwick J. Find Out How Much Traffic a Website Gets: 3 Ways Compared [Internet]. SEO Blog by
Ahrefs. 2018 [cited 2022 Feb 2]. Available from: https://ahrefs.com/blog/website-traffic/
88. Kashuba Margaret. SimilarWeb vs. SEMrush: Which Offers More Accurate Data? | CustomerThink
[Internet]. 2020 [cited 2022 Feb 2]. Available from: https://customerthink.com/similarweb-vs-semrush-
which-offers-more-accurate-data/
89. Pace Richard D. SimilarWeb: Fuzzier and Warmer than Alexa—Everything PR [Internet]. Everything
PR News. 2013 [cited 2022 Feb 2]. Available from: https://everything-pr.com/similarweb-alexa/
90. Times Internet. Insights—competition benchmarking tools in the internet industry | Times Internet
[Internet]. 2020 [cited 2022 Feb 2]. Available from: https://timesinternet.in/advertise/marketing/
insights/competition-benchmarking-tools-in-the-internet-industry/
91. Sahnoun Yassir. SimilarWeb Review: Know Your Audience, Win Your Market [Internet]. Monitor Back-
links Blog. 2018 [cited 2022 Feb 2]. Available from: https://monitorbacklinks.com/blog/content-
marketer/similarweb-review
92. Hogan Bruce. How to Analyze Competitor Website Traffic [Internet]. SoftwarePundit. 2020 [cited 2022
Feb 2]. Available from: https://www.softwarepundit.com/seo/competitor-taffic-analysis
93. Fishkin Rand. The Traffic Prediction Accuracy of 12 Metrics from Compete, Alexa, SimilarWeb, &
More [Internet]. SparkToro. 2015 [cited 2022 Feb 2]. Available from: https://sparktoro.com/blog/traffic-
prediction-accuracy-12-metrics-compete-alexa-similarweb/
94. Langridge Patrick. How Accurate Are Website Traffic Estimators? [Internet]. 2016 [cited 2022 Feb 2].
Available from: https://www.screamingfrog.co.uk/how-accurate-are-website-traffic-estimators/
95. Aguiar João. SimilarWeb vs SEMrush—Comparing Website Traffic (2020 Case Study) [Internet].
Mobidea Academy. 2020 [cited 2022 Feb 2]. Available from: https://www.mobidea.com/academy/
similarweb-vs-semrush-website-traffic/
96. Siegel DA. The mystique of numbers: belief in quantitative approaches to segmentation and persona
development. In: CHI ‘10 Extended Abstracts on Human Factors in Computing Systems. New York,
NY, USA: Association for Computing Machinery; 2010. p. 4721–32. (CHI EA ‘10).
PLOS ONE Measuring user interactions with websites
PLOS ONE | https://doi.org/10.1371/journal.pone.0268212 May 27, 2022 24 / 27
97. W3Techs. Usage Statistics and Market Share of Google Analytics for Websites, October 2020 [Inter-
net]. 2020 [cited 2022 Feb 2]. Available from: https://w3techs.com/technologies/details/ta-
googleanalytics
98. Google. Google Analytics Set up the Analytics global site tag—Analytics Help [Internet]. 2020 [cited
2022 Feb 2]. Available from: https://support.google.com/analytics/answer/1008080?hl=en
99. Google. Google Analytics About data sampling—Analytics Help [Internet]. 2020 [cited 2022 Feb 2].
Available from: https://support.google.com/analytics/answer/2637192?hl=en
100. Weber Jonathan. How Accurate Is Sampling In Google Analytics? | Bounteous [Internet]. 2016 [cited
2022 Feb 2]. Available from: https://www.bounteous.com/insights/2016/03/03/how-accurate-
sampling-google-analytics/
101. Salkind N. Triangulation. In: Encyclopedia of Research Design. 2455 Teller Road, Thousand Oaks
California 91320 United States: SAGE Publications, Inc.; 2010.
102. Jick TD. Mixing Qualitative and Quantitative Methods: Triangulation in Action. Administrative Science
Quarterly. 1979; 24(4):602–11.
103. SimilarWeb. SimilarWeb Marketing Solution: The Most Reliable and Comprehensive Data on Compet-
itor and Market Strategies. 2018.
104. SimilarWeb. What does connecting my Google Analytics account with SimilarWeb mean?–Knowledge
Center—SimilarWeb [Internet]. 2022 [cited 2022 Feb 2]. Available from: https://support.similarweb.
com/hc/en-us/articles/208420125-What-does-connecting-my-Google-Analytics-account-with-
SimilarWeb-mean-
105. Wasserman Yossi. How SimilarWeb analyze hundreds of terabytes of data every month with Amazon
Athena and Upsolver [Internet]. Amazon Web Services. 2018 [cited 2022 Feb 2]. Available from:
https://aws.amazon.com/blogs/big-data/how-similarweb-analyze-hundreds-of-terabytes-of-data-
every-month-with-amazon-athena-and-upsolver/
106. Lin J, Kolcz A. Large-scale machine learning at twitter. In: Proceedings of the 2012 international con-
ference on Management of Data—SIGMOD ‘12. Scottsdale, Arizona, USA: ACM Press; 2012. p. 793.
107. Nie Zaiqing, Kambhampati S, Nambiar U. Effectively mining and using coverage and overlap statistics
for data integration. IEEE Transactions on Knowledge and Data Engineering. 2005 May; 17(5):638–
51.
108. SimilarWeb. Connecting Google Analytics and Similarweb [Internet]. Similarweb Knowledge Center.
2022 [cited 2022 Jan 30]. Available from: https://support.similarweb.com/hc/en-us/articles/
208420125-Connecting-Google-Analytics-and-Similarweb
109. Boos DD, Hughes-Oliver JM. How Large Does n Have to be for Z and t Intervals? The American Statis-
tician. 2000; 54(2):121–8.
110. Box GEP, Cox DR. An Analysis of Transformations. Journal of the Royal Statistical Society Series B
(Methodological). 1964; 26(2):211–52.
111. Box GEP, Andersen SL. Permutation Theory in the Derivation of Robust Criteria and the Study of
Departures from Assumption. Journal of the Royal Statistical Society Series B (Methodological). 1955;
17(1):1–34.
112. Hull D. Using statistical testing in the evaluation of retrieval experiments. In: Proceedings of the 16th
annual international ACM SIGIR conference on Research and development in information retrieval.
New York, NY, USA: Association for Computing Machinery; 1993. p. 329–38. (SIGIR ‘93).
113. Avinash Kaushik. Web Analytics Standards: 26 New Metrics Definitions [Internet]. Occam’s Razor by
Avinash Kaushik. 2007 [cited 2022 Feb 2]. Available from: https://www.kaushik.net/avinash/web-
analytics-standards-26-new-metrics-definitions/
114. Kaushik A. Web Analytics: An Hour a Day. 1st Edition. Indianapolis, Ind: Sybex; 2007. 480 p.
115. SimilarWeb. SimilarWeb Category [Internet]. Knowledge Center—SimilarWeb. 2022 [cited 2022 Feb
2]. Available from: https://support.similarweb.com/hc/en-us/articles/360000810469
116. SimilarWeb. SimilarWeb All Categories [Internet]. SimilarWeb.com. 2022 [cited 2022 Feb 2]. Available
from: https://www.similarweb.com/category/
117. Bovbjerg ML. Random Error. In: Foundations of Epidemiology [Internet]. Oregon State University;
2020 [cited 2020 Oct 9]. Available from: https://open.oregonstate.education/epidemiology/chapter/
random-error/
118. Ramadan Alex. Common Google Analytics Setup Errors and Omissions [Internet]. UpBuild. 2019
[cited 2022 Feb 2]. Available from: https://www.upbuild.io/blog/common-google-analytics-setup-
errors/
PLOS ONE Measuring user interactions with websites
PLOS ONE | https://doi.org/10.1371/journal.pone.0268212 May 27, 2022 25 / 27
119. Gant Amanda. Inaccurate Google Analytics—Why Google Analytics is Wrong and How to Fix It [Inter-
net]. Orbit Media Studios. 2020 [cited 2022 Feb 2]. Available from: https://www.orbitmedia.com/blog/
inaccurate-google-analytics-traffic-sources/
120. Bloom Kevin. Why Your Google Analytics Data is Wrong and How To Fix It [Internet]. Hinge Marketing.
2020 [cited 2022 Feb 2]. Available from: https://hingemarketing.com/blog/story/why-your-google-
analytics-data-is-wrong-and-how-to-fix-it
121. Upton E. 88% of Shopify stores have Google Analytics set up incorrectly [Internet]. Econsultancy.
2018 [cited 2022 Feb 2]. Available from: https://econsultancy.com/shopify-stores-google-analytics-
set-up-incorrectly/
122. Williamson K, Burstein F, McKemmish S. Chapter 2—The two major traditions of research. In: William-
son K, Bow A, Burstein F, Darke P, Harvey R, Johanson G, et al., editors. Research Methods for Stu-
dents, Academics and Professionals (Second Edition). Chandos Publishing; 2002. p. 25–47. (Topics
in Australasian Library and Information Studies).
123. Google. Google Analytics Bounce rate—Analytics Help [Internet]. 2020 [cited 2022 Feb 2]. Available
from: https://support.google.com/analytics/answer/1009409?hl=en
124. Schneider Daniel, Trucks Ruth M. Bounce Rate: What you need to know and how to improve [Inter-
net]. SimilarWeb. 2021 [cited 2022 Feb 2]. Available from: https://www.similarweb.com/corp/blog/
bounce-rate/
125. Eraslan S, Yesilada Y, Harper S. “The Best of Both Worlds!”: Integration of Web Page and Eye Track-
ing Data Driven Approaches for Automatic AOI Detection. ACM Transactions on the Web. 2020; 14
(1):1:1–1:31.
126. Jiang T, Guo Q, Chen S, Yang J. What prompts users to click on news headlines? Evidence from
unobtrusive data analysis. Aslib Journal of Information Management. 2019 Jan 1; 72(1):49–66.
127. Vranica S. A “Crisis” in Online Ads: One-Third of Traffic Is Bogus. Wall Street Journal [Internet]. 2014
Mar 23 [cited 2022 Feb 2]; Available from: https://online.wsj.com/article/
SB10001424052702304026304579453253860786362.html
128. Acimovic J, Erize F, Hu K, Thomas DJ, Mieghem JAV. Product Life Cycle Data Set: Raw and Cleaned
Data of Weekly Orders for Personal Computers. Manufacturing & Service Operations Management.
2018; 21(1):171–6.
129. Sarokin David. What Is the Life Span of the Average PC? | Small Business—Chron.com [Internet].
2020 [cited 2022 Feb 2]. Available from: https://smallbusiness.chron.com/life-span-average-pc-
69823.html
130. Arkko J. The influence of internet architecture on centralised versus distributed internet services. Jour-
nal of Cyber Policy. 2020 Jan 2; 5(1):30–45.
131. Liu Shanhong. Desktop internet browser market share 2015–2020 [Internet]. Statista. [cited 2022 Feb
2]. Available from: https://www.statista.com/statistics/544400/market-share-of-internet-browsers-
desktop/
132. Abraham DM, Meierhoefer C, Lipsman A. The Impact of Cookie Deletion on the Accuracy of Site-
Server and Ad-Server Metrics: An Empirical Comscore Study. 2007;19.
133. Wills CE, Zeljkovic M. A personalized approach to web privacy: awareness, attitudes and actions.
Information Management & Computer Security. 2011 Mar 22; 19(1):53–73.
134. Opentracker. Third-Party Cookies vs First-Party Cookies [Internet]. Opentracker. [cited 2022 Feb 2].
Available from: https://www.opentracker.net/article/third-party-cookies-vs-first-party-cookies-2/
135. WRAL. A study of Internet users’ cookie and javascript settings [Internet]. smorgasbork. 2009 [cited
2022 Feb 2]. Available from: http://www.smorgasbork.com/2009/04/29/a-study-of-internet-users-
cookie-and-javascript-settings/
136. Google. Google Browse in private—Computer—Google Chrome Help [Internet]. 2020 [cited 2022 Feb
2]. Available from: https://support.google.com/chrome/answer/95464?co=GENIE.Platform%
3DDesktop&hl=en
137. Habib H, Colnago J, Gopalakrishnan V, Pearman S, Thomas J, Acquisti A, et al. Away From Prying
Eyes: Analyzing Usage and Understanding of Private Browsing. In: Fourteenth Symposium on Usable
Privacy and Security. Baltimore, MD, USA; 2018. p. 18.
138. Fettman Eric. A Sweet Treat, But Users Delete: Cookies and Cookie Deletion in Google Analytics
[Internet]. Cardinal Path. 2015 [cited 2022 Feb 2]. Available from: https://www.cardinalpath.com/blog/
cookies-and-cookie-deletion-in-google-analytics
139. Ringvee S. How to Detect and Track Incognito Users with Google Analytics and Google Tag Manager
[Internet]. Reflective Data. 2019 [cited 2022 Feb 2]. Available from: https://reflectivedata.com/how-to-
detect-track-incognito-users-with-google-analytics-and-google-tag-manager/
PLOS ONE Measuring user interactions with websites
PLOS ONE | https://doi.org/10.1371/journal.pone.0268212 May 27, 2022 26 / 27
140. Pew Research Center. Demographics of Mobile Device Ownership and Adoption in the United States
[Internet]. Pew Research Center: Internet, Science & Tech. 2019 [cited 2022 Feb 2]. Available from:
https://www.pewresearch.org/internet/fact-sheet/mobile/
141. Spangler Todd. U.S. Households Have Average of 11 Devices. 5G Will Push That Higher—Variety
[Internet]. 2019 [cited 2022 Feb 2]. Available from: https://variety.com/2019/digital/news/u-s-
households-have-an-average-of-11-connected-devices-and-5g-should-push-that-even-higher-
1203431225/
142. Matomo. Matomo—The Google Analytics alternative that protects your data [Internet]. Analytics Plat-
form—Matomo. 2020 [cited 2022 Feb 2]. Available from: https://matomo.org/
PLOS ONE Measuring user interactions with websites
PLOS ONE | https://doi.org/10.1371/journal.pone.0268212 May 27, 2022 27 / 27