In the week where Universal Analytics has been sunset by Google. It seemed like a good opportunity to outline how Universal Analytics has changed the way transactional websites captured and reported on performance. This post will outline, what I think are, the 5 ways Universal Analytics improved web performance analysis. And how this enabled businesses to improve their understanding of where users were coming from, who the users were and how they behaved.
What makes it possible for Universal Analytics to capture data?
If you work for a company with a transactional website the chances are you will be somewhat familiar with what Google Analytics. But perhaps have wondered how it does what it does. The answer is tracking.
Within the JavaScript, of the website, there will be a snippet that sends data, such as every new page load, to Google Analytics. Overtime the data that gets sent, via the JavaScript snippet, has become more complex. With changes being made to increase the amount and type of data that gets sent.
Why is it important to capture performance data?
It is important to capture data of transactional websites to understand if the website is functioning as expected. This is true for both a sales and usability perspective. Though if your website it throwing up 404s and conducting an endless loading loop it will be pretty obvious to spot that it is not working as expected.
The Google Analytics that we know and love (or loathe) today, has been through several iterative changes. In fact, Google acquired Urchin (its birth name let’s say) in 2005 from Quantified Systems aka Urchin Software Corporation. Which developed the tool to track the number of visits to websites so that they could bill for bandwidth.
Changes made since then have made it possible to see where visitors are coming from. How they are interact with different parts of the website and flag user journey errors that don’t necessarily result in a 404 error, or the like.
5 ways Universal Analytics improved performance analysis of transactional websites
1. Comparing performance between device types
Arguably the biggest impact Universal Analytics had, on the analysis of web performance, was the ability to collect data from any device.
The tracking tools that came before Universal Analytics, were all built on the foundations of Urchin. Which primarily involved the use of several cookies, and gif image files, (yep, that surprised me as well) to keep track of sessions and the onsite user behaviour of visits.
Universal Analytics changed this by tracking sessions, and on site user behaviour, through a single cookie. The cookie would assign a non-personally identifying visitor ID to a visitor’s browser. Making it possible to collect gradual detail such as browser and device type.
The introduction of a new Measurement Protocol, to support the single cookie tracking. Made it possible to assign a cookie to any device with the capability of connecting to the internet. Making it possible to see the differences between user behaviour, when a website was accessed via a variety of devices. Such as desktop computer or laptop, portable device or games console.
2. Incentivising users to login
As businesses gained visibility of the behavioural, and performance, differences between devices. The next question to answer was, how does this relate to the consumer conversion funnel? Enter, userID.
Universal Analytics made it possible to track the same user across multiple devices, as long as there was a unique identifier to link each of the cookies placed. Something like a login. The login could then generate a userID – ensuring that this was still non-personally identifiable. With the UserID dimension, reports could then be built to analyse how UserIDs would work through the conversion funnel between devices. Perhaps spending a longer duration per page on mobile, in a research phase. And spending considerably less time on desktop with a higher conversion propensity.
3. Web customers are not one homogeneous group
With the increasing customer-centric approach to the data made available in Universal Analytics. Understanding the differences between different user profiles became more important, as it could be used to inform the actions needed to incentivise behaviours most desirable to business goals.
It became possible create segments based on customer attributes. Surfacing differences in behaviour based on how a customer arrived on a website or whether they were logged in, for example. Custom segments made it possible to filter data to focus on specific customer groupings. To help inform buyer personas so that parts of the site, and products offered, could be more geared towards those customers.
4. Behaviour and conversion related KPIs
Along with custom segments, came custom dimensions and custom metrics.
In Universal Analytics a dimension is categorised as an attribute of data and a metric is the assessment measurement. For example, if the ask was to review the visits by country, the metric would be visits and the dimension would be the countries.
Custom dimensions and metrics make it possible to collect and analyse data that is not automatically tracked. For dimensions this could be attributes for logged-in and guest customers, or customer attributes from stored information in a CRM system. Custom metrics are useful in assessing how customers behave once they are on site. For example within how many page views to they view a PDP from an internal search result.
Custom metrics made it possible to develop measurable KPIs related to user behaviour. As a way to assess the impact of improvements made to customer journey pain points.
5. Visibility of traffic driving Search terms
Google is the biggest search engine and a key gateway to acquiring new customers. Having visibility of which search terms are directing traffic to your site, can help to assess how well pages are ranking for business specific keywords.
Granted there are specialist Search Engine Optimisation (SEO) tools out there, that have the capabilities to give more insights. Having visibility of traffic driving search terms and the behaviour of the users from that result will help to show if there is more work needing to be done on site to improve the user experience. If the results on a page don’t match the expectation of the search query, for example, this could lead to a high bounce rate.
Many of the features we enjoyed in Universal Analytics came as a result of a change in business needs. Businesses “globalised online”, making a need to assess how users from different geographical locations interacted with a website. Advances in technology made websites more accessible, making a need to assess how performant and usable websites were between devices.
With the introduction of GA4, new features will be introduced once again to address changes in business needs. From technical changes in how onsite behaviour is recorded. To practical changes like bringing performance analytics from different digital platforms into one place. As with all change there will be teething problems transitioning over the new platform. But I look forward to seeing what improvements are instore for performance analysis and user experience.
Like this post? You should see my newsletter.
Subscribe to the Excel & ME newsletter to get updates that will empower you to inform your business decisions with data.
