This publication is broken up into three sections:
TL;DR - For those wanting a quick take
Summary - For those wanting a bit more context and high level points
Article - Main body of work containing fully detailed article and explanations that you might want to consume over several readings
TL;DR
In the early days of the web (1993), websites were “web pages” and everything was static.
The next major evolution (1997) in web analytics came manual tagging.
In 2005, there was no SaaS category. Enter Google Analytics, as a tool to measure site traffic, and in turn to assess the effectiveness of a company’s marketing campaigns, GA was unprecedented.
From 2008 onwards tools like KISSmetrics, and Mixpanel, let you see what your users are doing and develop hypotheses about why they are behaving a certain way.
Today’s web environment is much different. It is no longer a static page with very few interactive elements. It’s dynamic, with the rise of A/B testing and experimentation to personalization.
With the 4th innovation in web analytics, technology can be leveraged to capture all web data for you automatically. Technology like autocapture also organizes all the data for you in a visualization layer, freeing you from tedious tagging and allowing you to spend time slicing and visualizing it any way you need.
Summary
In the early days of the web (1993), websites were “web pages” and everything was static. These web pages didn’t require much effort to track, so the hit counter was all that was needed. Anytime a web page element was requested or loaded it was called a “hit,” a very basic metric for recording page visits.
The next major evolution (1997) in web analytics came manual tagging. The web wasn’t dynamic just yet, so the page you loaded was the page you would be interacting with. This was perfect for manual tagging. There weren’t many elements on the page, pages were not complex, and when you tagged your website or product, nothing else needed to change unless you redesigned your interface.
In 2005, there was no SaaS category. Software was still installed via CD-ROM aka CDs for short. Enter Google Analytics. As a tool to measure site traffic, and in turn to assess the effectiveness of a company’s marketing campaigns, GA was unprecedented. In time, its capabilities grew, and GA is still a powerful tool for calculating ROI for advertising spend.
From 2008 onwards tools like KISSmetrics, and Mixpanel, let you see what your users are doing and develop hypotheses about why they are behaving a certain way.
All of the web analytics tools mentioned so far are built on the philosophy of tagging events. Tag managers help with this problem, but they’re only a band-aid on a fundamentally outdated philosophy when it comes to web data collection and data infrastructure. These tools are rooted in the approach that web data collection and analytics needs to have implementation and tagging as the first step.
Today’s web environment is much different. It is no longer a static page with very few interactive elements. It’s dynamic, with the rise of A/B testing and experimentation to personalization. Web analytics tools based on event tagging encounter limitations in today’s dynamic environment.
With the 4th innovation in web analytics, technology can be leveraged to capture all web data for you automatically. Technology like autocapture also organizes all the data for you in a visualization layer, freeing you from tedious tagging and allowing you to spend time slicing and visualizing it any way you need.
Often, the answers to the questions we ask lead us to new questions. If you need data on something you didn’t think to tag six months ago, the web analytics technology should have that available retroactively at your fingertips.
Today, web analytics tools have the capability to track individual user behaviour, analyse user behaviour patterns, and provide insights into how users interact with websites.
Despite the shortcomings of pure web analytic solutions like GA they still have an important place and role to play in your analytics set-up.
Article
Web analytics has been around since the early days of the World Wide Web. Web analytics in its most basic form is a way of creating insights from tracking key interactions taking place on web pages.
Web analytics includes the process of analyzing, collecting, and reporting data about the behavior of visitors to a website. It helps website owners to understand how visitors interact with their website and make data-informed decisions to improve their website's performance and user experience.
Since 1993, the Internet has seen several significant changes in the way web traffic is measured. Web analytics has changed from humble beginnings as hit counters for web pages, to server log events, to manual event tagging, and now we have automated technology in the form of auto-capture that tracks all customer interactions on the web page interface.
Despite web analytics being an industry segment worth over $3bn, brands still say they don’t have the tools they need. There is plenty of room for growth and expansion in the analytics market.
Historical overview of web analytics
Log file data in 1993
In the early days of the web, websites were “web pages” and everything was static. These web pages didn’t require much effort to track, so the hit counter was all that was needed. Anytime a web page element was requested or loaded it was called a “hit,” a very basic metric for recording page visits.
In the very beginning, a basic understanding of web page activity could be grasped by reading and interpreting server log file data. Server log files kept a record of HTTP request information with attributes such as the source IP address, timestamp, the request type, status, and bytes transferred. But the problem with this data is that it required technical skills to access and interpret.
Around this time, web page counters were popularized, with the most popular being HitWebCounter. These counters gave you a basic idea of your product and company’s health but zero visibility into what your users were doing on your website.
Then came Analog, the world’s first free log file interpreter, which paved the way for the analytics industry. Some enterprising technology people (Stephen Turner and Christopher Tilley) created Analog which was the first free log file analysis tool which made log file data more understandable to the average user. Only premium analytics solutions like Business Objects (acquired by SAP) helped users understand the insights from these technical sources, and they were out of the price range for most people and organisations except enterprise businesses.
Manual tagging and pageviews in 1997
The next major evolution in web analytics came manual tagging. The web wasn’t dynamic just yet, so the page you loaded was the page you would be interacting with. This was perfect for manual tagging. There weren’t many elements on the page, pages were not complex, and when you tagged your website or product, nothing else needed to change unless you redesigned your interface.
Urchin was founded in 1995 and Omniture in 1996 which were precursors to Google Analytics. Urchin was later acquired by Google and was incorporated into Google Analytics circa 2005. Both products allowed for the creation of dashboards, reports, and calculated metrics. The legacy of Urchin lives on in acronym UTM (i.e., see below story for more).
Back-story on Urchin from Tracking Garden
“Urchin was a web analytics software developed by Urchin Software Corporation. The original company was founded in 1995 by Paul Muret and Scott Crosby. Urchin was originally a software to evaluate web server log files, but then developed into a good web analysis tool (for those times, maybe in some points even for today).
In April 2005, Urchin (at that time in Version 6) was taken over by Google. After a last release (which was called “Urchin 7 by Google”) the official Urchin history ended and the way of Google Analytics began. In fact, parts of Urchin should be felt in Google Analytics for many years to come.
Some things in Google Analytics are still reminiscent of Urchin today. Up until Google Universal Analytics, a Analytics Property IDs began with UA - this was the abbreviation for Urchin Analytics. Or you surely know UTM parameters. Did you know that UTM stands for "Urchin Traffic Monitor"? And so there are still a few more points that remind us of Urchin in Google Analytics to this day.”
You now had visibility into things like pageviews, sessions, and other things that today we would consider to be fundamental web metrics. You could even filter by date ranges. All of this gave you a basic understanding of your company’s web site health.
However, you still had very little visibility into what your users and cohorts of users were doing. Tags were difficult to implement manually, and as the web became more dynamic, manual tagging became more painful to maintain.
Implementation became a sore point for most medium to large businesses because of the constant need to add, remove, and fix tags. This led to the development in the industry of tag implementation consultants who worked alongside businesses to plan, implement, and maintain their analytics platforms.
Google Analytics circa 2005 as the prototypical web analytic product of the time
In 2005, there was no SaaS category. Software was still installed via CD-ROM aka CDs for short. Business software was entirely on-premise, and even huge enterprise companies waited for yearly updates from the software maker.
At that point, websites were simple. Most contained a few static pages built with HTML and CSS. HTML 4 was eight years old. Its replacement, HTML 5, was nine years away. Ruby on Rails and Django were both in version 1.0. The first modern front-end frameworks, backbone.js and angular.js, wouldn’t show up for another five years.
Given that most websites were simple and static, the information that mattered to most companies was:
How many people came to your site?
How did those people or bots get there?
If a team could use this knowledge to increase the number of people who came to their site, they were gold.
Enter Google Analytics. As a tool to measure site traffic, and in turn to assess the effectiveness of a company’s marketing campaigns, GA was unprecedented. In time, its capabilities grew, and GA is still a powerful tool for calculating ROI for advertising spend.
Overview of Google Analytics
Google Analytics is a free web analytics tool that tracks and reports website traffic.
It helps you measure how users interact with your website content and provides insights into their behavior.
It can be used to track website performance and to analyze user engagement.
It provides detailed reports and data visualizations to help you better understand your website visitors.
You can use it to track goals, conversions, and ecommerce activities.
It integrates with Google AdWords, so you can see how your ads are performing.
You can use it to create custom dashboards and alerts to monitor the performance of your website.
You can also use it to track user interactions across multiple devices.
It offers advanced features such as funnel visualization and user segmentation.
It can be used to track almost any type of web activity.
User-centric behaviour in 2008
By now more complicated questions like,”What was the drop off between step 3 and step 4 of our 5-part funnel, and how can we improve that?” were being asked. The unmet needs of this period allowed data-driven, growth-focused companies to flourish and gave rise to an entire industry of conversion rate optimizers (CROs), funnel hackers, and growth marketers.
The people with these jobs were the primary drivers in the shift from session-based web analytics tools like Google Analytics or Adobe Analytics to user- centric web analytics tools. These web analytics providers still require manual tagging, but fancier visualizations were added to make things more human readable. Tools like KISSmetrics, and Mixpanel, let you see what your users are doing and develop hypotheses about why they are behaving a certain way.
Analytics also became more approachable for the non-technical marketer and business teams due to better visualizations. Using these tools, you could create funnels and cohorts to monitor users throughout their lifecycle. It could even work for mobile web and native iOS and Android.
Beyond visualizations, the most promising improvement in analytics during this period was the ability to dive deep into customer behaviors. This user-centric approach was fundamentally different than the previous web analytics tools. Now marketers, analysts, and product managers can understand what their users were doing and why.
Even with these massive strides in the web analytics space, all the legacy tools mentioned are still built on early 90’s, static web architecture: manual tagging. The most significant pain point is implementation, tagging, and continued maintenance of event tagging.
2013 – Automatic data capture and retroactive analytics
All of the web analytics tools mentioned so far are built on the philosophy of tagging events. Tag managers help with this problem, but they’re only a band-aid on a fundamentally outdated philosophy when it comes to web data collection and data infrastructure. These tools are rooted in the approach that web data collection and analytics needs to have implementation and tagging as the first step.
Today’s web environment is much different. It is no longer a static page with very few interactive elements. It’s dynamic, with the rise of A/B testing and experimentation to personalization. Web analytics tools based on event tagging encounter limitations in today’s dynamic environment.
With the 4th innovation in web analytics, technology can be leveraged to capture all web data for you automatically. Technology like autocapture also organizes all the data for you in a visualization layer, freeing you from tedious tagging and allowing you to spend time slicing and visualizing it any way you need. Often, the answers to the questions we ask lead us to new questions. If you need data on something you didn’t think to tag six months ago, the web analytics technology should have that available retroactively at your fingertips.
Your web analytics technology should enable you to have access to any click, swipe, tap, form change, or other events that occur on your site or app. This technology should free you from spending time tagging events, and instead spend time finding insights and business- critical data points.
This innovation causes a massive shift in thinking. Instead of being restricted to asking pre-defined questions instead you can dynamically generate and think of new questions as new data insights arise.
What can we measure with web analytics?
Web analytics started off as a simple way to track the number of visitors to a website, the number of page views and the time spent by visitors on a page.
Today, web analytics tools have the capability to track individual user behaviour, analyse user behaviour patterns, and provide insights into how users interact with websites.
Web analytics has evolved and become much more sophisticated. With web analytics, you can measure various aspects of website performance such as:
page views,
visitors,
traffic sources,
time spent on pages,
bounce rates,
conversions,
user engagement,
user behavior, and
user journey
What are the limitations of web analytics?
While tracking inbound sources is still important, great products are now built from understanding how users behave on your site or in your app.
What do they click on most often? Which features do they use? What paths do different users take? What behaviours correlate with retention? With conversion? What actions do people tend to do right before an event, and where do they get stuck?
It’s by answering questions like these that product teams can make the most impact in their product. The problem is that GA isn’t built to answer them. Yes, it’s possible with an enormous amount of work to force GA into copy of a tool that does this natively, but even then, you still won’t get all the data you need.
With the increased adoption of eCommerce and Fintech services, web apps are now as or more important than brick- and-mortar stores. More powerfully, we’ve seen software overwhelmingly move to SaaS, which literally did not exist when GA was founded.
The main limitation of web analytics is that it can only measure what is happening on the website. It cannot track the user’s behavior outside the website, in the physical world.
Additionally, web analytics cannot track user activity in real time, as it takes some time for data to be collected and analyzed.
Also, web analytics does not provide in-depth insights into user behavior or preferences.
Web analytics tools can be inaccurate if not configured properly, as they may not include all the necessary data points and metrics.
Legacy web analytic tools have significant pain points around creating tracking plans, implementation, and continued maintenance of event tagging; and if you want to do serious data analysis in a BI tool, many experience major difficulties with the underlying data including inconvenient schemas, missing data, and having to manipulate data.
Finally, web analytics is not able to measure the effectiveness of a website’s design and user experience other tools map be needed like session replay technology or page heatmap visualisation.
Recommendations for actionable web analytic insights
Despite the shortcomings of pure web analytic solutions like GA they still have an important place and role to play in your analytics set-up.
Below are some ideas on how best to leverage tools like GA for now.
Goals and objectives: Define the goals and objectives of your analytics to determine how to measure success.
Data Collection: Collect data from all sources, including website, campaigns, and customer interactions.
Site Analysis: Analyze site performance, visitor behavior, and user journeys to identify areas for improvement.
Campaign Tracking: Track campaigns to measure their effectiveness and identify the most successful strategies.
Attribution Modeling: Use attribution modeling to identify which channels and campaigns are most effective.
Optimization: Optimize website and campaigns for maximum performance by experimenting with various elements.
Segmentation: Segment visitors and customers by demographics, interests, and behavior to target audiences.
Analysis Tools: Utilize analytics tools such as Google Analytics, Adobe Analytics, and Tableau to gain insights.
Reporting and Dashboards: Create reports and dashboards to track KPIs and measure performance over time.
Insights: Leverage insights to inform decisions, create strategies, and improve user experience
Despite the shortcomings of pure web analytic solutions like GA they still have an important place and role to play in your analytics set-up.
Post Script
Before you go, please could you do the following?
Subscribe
Share
Survey
Star
If you got value from reading the article a star liking would be highly appreciated!
Resources
Analog - https://en.wikipedia.org/wiki/Analog_(program)
AWStats - https://en.wikipedia.org/wiki/AWStats
Heap – How to Increase Revenue by Measuring Customer Behaviour