Web analytics
Web analytics is a system for measuring, collecting, analyzing, presenting and interpreting information about website visitors in order to improve and optimize them. The main task of web analytics is monitoring website traffic, based on which the site’s audience is determined and visitors’ behavior is studied to make decisions on the development and expansion of the web resource’s functionality. Web analytics allows you not only to work on improving sites, but also to work on optimizing the budget for online promotion.
In 1990, with the birth of the HTTP protocol, the era of web analytics began. It became possible to record user and server interactions in log files. Every time an Internet user calls an HTML element, a hit string is written to the log file. With an increase in traffic, site owners began to receive too many hits, log files increased in size. What was needed was a way to analyze such volumes of data.
The beginning of commercial web analytics can be considered the creation of WebTrends in 1993
In 1995, the Analog system was created - the first free system for analyzing log files. Analog allowed generating reports from log files, had clear documentation and the ability to graphically interpret data. Finally, not only programmers, but also professional marketers were able to do web analytics.
Over time, pages began to contain a larger number of elements - loading an element ceased to mean loading the page. There was a new method of collecting information about visiting pages - javascript tags, which became the most common with the development of networks. Javascript tags are implemented by the owner on all pages of the site. When the page loads, a tag is executed - it collects information about the user's visit and saves it to the database.
In 2005, Google created its Google Analytics web analytics system by acquiring Urchin Software Corporation and its Urchin project.
In 2006, an analysis of the behavior of visitors on the page started. A record of visitor behavior on the page began to be kept. Clicks and scrolling cards are formed. Web analysis has ceased to be only quantitative, but has also become qualitative.
Application area
Web analytics helps in many aspects of website development and online promotion. Here are the main ones:
Development of site functionality based on trends in visitors behavior
Evaluation of the effectiveness of advertising campaigns and search engine promotion on the Internet
Identification of problem areas in the structure, navigation and content of the site
Optimization of the product line presented on the site
Statistics of traffic to sections and web pages of the site allows you to understand:
number of web pages viewed
keywords and phrases by which visitors find the site in search engines,
geography of visitors
time spent on a web page by a visitor,
Navigation between web pages
site audience (casual, regular visitors, etc.)
Convenience of site navigation for visitors, etc.
Web analytics methods
Analysis of site traffic: statistics, trends, absolute and relative indicators
Analysis of data from electronic commerce: average bill, popular products, revenue by traffic attraction channels
Usability analysis: analysis of the density of clicks, conversion paths of visitors to the site, analysis of scrolling
Analysis of the behavior of visitors on the page: interaction with forms, micro and macro conversions
Benchmarking. Comparison with general trends and competitors using independent platforms (Alexa, GemiusAudience, Google Trends)
End-to-end analytics. Tracking the user's full path from viewing ads to completing a transaction, as well as repeat sales.
Cookies collection. This allows analytical services to correlate user activity on web resources where he provided personal information. Further, based on the collected data about users, targeted advertising can be configured for them. However, confidentiality concerns regarding cookies have led to a noticeable minority of users blocking or deleting third-party cookies. In 2005, some reports showed that about 28% of Internet users blocked third-party cookies, and 22% deleted them at least once a month.
Basic Web Analytics Terms
Today, there are still no globally agreed definitions for terms related to web analytics. The main organizations that contributed to this area were IAB (Interactive Advertising Bureau), JICWEBS (Joint Industry Standards Committee in the UK and Ireland) and DAA (Digital Analytics Association), officially known as WAA. (Web Analytics Association, USA). However, many terms are actively used by them and therefore the following list may be a useful starting point:
The bounce rate is the percentage of visits to one page without any other action on that page, or a session in which there was only one request to the server.
Click path (English Click path) - a chronological sequence of page views within a visit or session.
Hit - a file request from a web server (for example, a web page, image, JavaScript or a cascading style sheet).
Pageview (eng. Pageview) - showing one page of a site, in other words, a request to download one HTML file (web page) of an Internet site. Displaying and viewing a page is often confused, but these are fundamentally different concepts. A single page impression can generate multiple views, since all image files, .js and .css are also requested from the web server.
Unique Visitor / Unique User (Unique Visitor / Unique User) - a client with a unique identification that generates page views or visits over a certain period of time (for example, day, week or month). Authentication is usually done with a persistent cookie, which was placed on the computer using the page code of the site. A “visitor” is not the same as a person sitting at a computer while visiting a resource, since an individual can use different computers or use different browsers on the same computer, and will be considered as a different visitor in each case. Increasingly, visitors are uniquely identified using the Flash LSO (Local Shared Object), which are less sensitive to data privacy.
Visit / Session (English Visit / Session) - a period of time during which the user is actively working with a website or application. [7] All data about the use of the site or application is attached to the session: page views, events, electronic commerce transactions, etc. A visit or session is defined as a sequence of page requests or, in the case of tags, image requests from the same uniquely identified client. A visit is considered completed if no requests have been recorded within a certain number of minutes. The 30-minute limit (“timeout”) is used by many analytic tools, but in some tools (such as Google Analytics) it can be changed to a different number of minutes. Analytics data collectors do not have a reliable way to find out if a visitor has been browsing other sites between page views; a visit is considered one visit until the events (pageviews, clicks and everything that is recorded) last 30 minutes, unless otherwise configured in analytics.
Active Time / Engagement Time - The average amount of time that visitors spend actually interacting with content on a web page is calculated based on mouse movements, clicks, freezes, and scrolling.
A click is an event that occurs when a user clicks on a control.
An event is a single action or chain of actions that occurs on a website. Pageview is an event type.
The bounce rate (English Bounce Rate) - was originally defined in the Google Analytics reports and is interpreted in the original as the percentage of visitors who viewed no more than 1 page for a session. statistics applied to a single page, not to a website, are measured as a percentage.
First Visit (eng. First Visit) - a visit to the site by a uniquely identified client who, theoretically, has not made any transitions to this web resource before. Since the only way to find out if a previously identified customer was on the site is to have a persistent cookie or digital fingerprint obtained from a previous visit, the “first visit” label is not reliable if the cookies were deleted from the site from the time of their previous visit.
Frequency (Engl. Frequency) shows the frequency of customer visits to a website in a certain period of time. It is calculated by dividing the total number of sessions (or visits) by the total number of unique visitors for a specified period of time, for example, a month or a year.
Impression is an ad that appears on the page you are viewing. It may appear on the page you’re viewing below the area actually displayed on the screen, so most impression metrics do not necessarily mean that the ad was visible.
Page viewing time (English Page Time Viewed) - the time during which a single page is displayed on the screen (or a blog, an advertising banner, etc.) is measured as the calculated difference between the request time for this page and the time of the next recorded request. If there is no next recorded request, then the viewing time of this page is not included in the reports.
A heat map (Site Overlay) is a report method in which statistics (clicks) or “hot spots” are superimposed by physical location on a visual snapshot of a web page, a heat map displays the activity of users on the site.
Web analytics tools
You can collect statistics using:
Counters are external programs. To get statistics, a small code fragment (usually 1-2 kilobytes) is installed on the website’s web pages. The point is that when you enter the site, the browser loads a picture that is posted on the information collection site. Data on counter downloads are entered into a database, which can be placed on the server of the provider of statistics collection and processing services, and then viewed, for example, on its website.
Using log analyzers, internal programs that collect data accumulated by the server.
Comparison of log analyzers and counters
Log analyzers allow you to collect statistics without changing anything on the site. The web server independently creates the log files and saves them to the server. Data is stored on the company's servers in a standard format. This allows companies to create their own programs for data analysis, to switch to updates exactly at the moment when they need it. The log files contain information about the behavior of search robots, which allows you to correctly evaluate the work on SEO optimization.
Counters count the opening of the page only after it loads. Due to this, they can take into account visits to cached pages, which is impossible with log analyzers. You can access not only standard information on visits, but also the number of purchases, clicks on certain buttons and the like. Companies that do not have servers can store information for web analytics if they use meters. Counters are currently the standard in web analytics.
Log Analyzers
Main article: Log Analyzer
WebTrends (WebTrends)
Webalizer
AWStats
Web analytics systems (obsolete name - tracker counters)
Summarize information on visits selected for a certain slice (measurements) specified by the user.
Google analytics
Piwik
Yandex.Metrica
Liveinternet
Рейтинг@Mail.ru
OpenStat (formerly Spylog)
Hotlog
Adobe analytics
Clicktale
Systems of Internet statistics with detail on page views
In addition to the summary information, they provide information on page views within each visit.
Woopra
Yandex.Metrica
Internet analytics systems detailing visitor behavior on the page
They give the greatest possible detail with the ability to view all the actions of visitors: mouse movements, clicks, keystrokes, etc. Based on the collected behavioral information, reports are constructed in the form of visitor activity maps on the page.
Yandex.Metrica
Spybox
Adobe analytics
Tag manager
It allows you to insert the code on the site only once, and carry out all other manipulations involving changing the code on the site within yourself. Simplifies the process of installing counters and other tracking elements (beacons, tracking pixels) on the site.
Google tag manager
Tealium
Adobe Dynamic Tag Management (DTM)
Signal
Qubit
Piwik
Tagcommander
Ensighten
Rating Counters
They give the number of visitors per day, week, month, for the whole history.
Rambler's Top100
Liveinternet
Рейтинг@Mail.ru
Openstat
Hotlog