Since 2020, Google has been transitioning its web analytics tracking system from Universal Analytics (UA) to Google Analytics 4 (GA4). Rather than primarily tracking website sessions and pageviews as UA did, GA4 tracks a range of more granular events. This method is intended to provide a deeper look at the user journey across websites and mobile apps as well as being more customizable for the needs of analytics users.

The legacy UA system officially stopped tracking data on July 1, 2023. Unfortunately, the transition to GA4 has not been a smooth transition for many. In addition to the basic logistics of switching tracking systems, users have reported difficulty finding the reports they had relied upon for years with UA, missing basic data dimensions, and a steep learning curve working with deeper data explorations.

For those who have come to rely on UA for data analysis, marketing strategies and business decisions, GA4 may feel like a step back. Whether or not this will remain the case as GA4 evolves, there is a silver lining in that these challenges can provide an opportunity to rethink our approach to analytics in general. Many have come to rely on UA as a single authoritative source of analytics data – but would multiple sources of data serve us better?

Analytics Data and Integrated Web Properties

Two decades ago when Google first acquired the platform that would become the foundation for Google Analytics, websites were more self-contained systems. This lent itself to looking at web traffic data as a unified source of intelligence.

Using a single source of data for analytics is in practice what many website owners have been doing with Google Analytics since then, and it is understandable why. UA was easy to set up and provided a wealth of data and reporting options. It was quick to get up to speed in exploring the data on the fly as well as to track key indicators through default reports. Event and conversion tracking customizations were straightforward to set up and allowed us to monitor key performance indicators (KPIs).

Today however, web properties regularly involve integrations between numerous separate systems ranging from content management platforms, applicant tracking systems, customer relationship managers and a slew of other communications and marketing tools. Each has its own method of collecting user data and reporting mechanisms. 

Continuing to rely on a single source among these systems is limiting and creates an illusion of thoroughness when we treat that single dataset as authoritative. Among the issues this can cause are: 

  • Basing strategic and resource decisions on incomplete data risks poor decision-making.
  • A sole dataset often exists in the purview of a single operational department or person, which can lead to the priorities and biases of that entity holding inordinate sway.
  • Errors in tracking or reporting do happen and without a point of comparison they may go undetected. 
  • KPIs often do not map to a specific data point in our analytics data model. Without the ability to combine datasets, we may not be able to shine a light on the factors contributing to our goals, or even the KPIs themselves.

The good news is that we don’t have to settle for a single dataset, or even multiple separate sources of data. Just as web properties themselves have shifted towards system integrations, our analytics approach should as well. This can be achieved both through data processing tools that help aggregate and analyze that data, as well as processes that bring together teams to regularly assess goals, analyze the data and iterate on strategy. 

Challenges on the Road to Data Source Integration

However, such integrations don’t come without their own challenges. After all, there’s a reason many have stuck with a single, simple source of analytics data all this time even as their systems have clearly expanded. 

Data Blending and Analyses Tools

Tools for combining data from separate sources are central to gaining more thorough and diverse insights. Ideally, such systems will be able to handle multiple steps of the data blending process including acquisition of the data from various sources, centralized storage, cleaning and standardization of the data set, creation of indicators from multiple data facets, reporting and visualization and ongoing updates and/or access to live data.

There are a lot of options for such systems – and GA4 hopes to be one of them. Regardless, the level of sophistication required to implement and maintain these is higher than the old UA-only method meaning it may require more staff expertise, setup time and ongoing maintenance than in the past. 

Standardized Reporting vs Data Exploration

An aspect of UA that was widely appreciated was the ability to explore live data on the fly to discover new data relationships and trends. This was in addition to the default overview data charts and tables. The combination allowed users to find the data visualizations that mattered most to them, while continuing to have the option to dig deeper as the need or want arose.

Nothing about using data from multiple sources needs to change that in theory. However, the varying datasets can make the processes more complex. Whereas UA would lead users to see certain indicators as meaningful, more forethought is required to make sense of the depth and breadth of these more complex arrays of data. And in long run this planning is a good thing! But it does require thinking about what data and trends are meaningful at the outset in order to make sure we are tracking data and combining in meaningful ways.

Obstacles to Data Sharing

A website owner typically has full access to determine if and how data is tracked on the website itself, and how those analytics will be reported and shared. When we start working with data sources from multiple integrated systems, we can run into challenges gaining access to the data we need. Sometimes those are technical issues – a platform may not track the data we need, or may not have an established means of sharing it with a third-party. Sometimes they are procedural issues – a team responsible for analytics on one platform may not appreciate the value of blending that data with related systems, or may actively be reluctant to share that data. 

The first steps to dealing with technical issues is to review documentation and have technical POCs consult with one another. Often there is an existing method to share data through an API or data export process, and even when not there may be a project roadmap that can be adjusted to prioritize creation of a solution. Procedural issues can sometimes be a tougher nut to crack, but when we establish a more integrated project team with a more agile approach to collaboration we can often identify shared goals that will help get that ball moving.

In addition, data privacy concerns are an increasing concern for data collection and sharing in general. 

Blended Web Analytics in the Real World

WHITE64 has helped clients in various industries move beyond Google Analytics as a single source of data for assessment of marketing campaigns and web property performance. Some examples include:.

  • Transportation Service Business: This client’s website accepts applications and allows users to manage monthly accounts through a proprietary account management system. Their customer care department sought to understand client needs more effectively and reduce the incidence of high-touch support requests. By injecting traffic source tracking into applications and support communications, we were able to blend UA data with account information and create an internal data reporting interface that the client used to identify areas for improvement both in the user interface as well as their team processes which ultimately led to more efficient customer care operations.
  • B2B Product/Service Provider: This client was looking to inform marketing campaign strategies across industry vertical sales teams. Their default UA data was used to gauge user interest in each topic area relative to each other, but not able to provide adequate insights within individual verticals. Sales teams also had varying levels of aptitude working within Google Analytics itself. We used Google Data Studio to blend data from UA, the CRM system, and marketing campaigns; develop vertical-specific reports and indicators; and automate distribution of reports to be utilized in the sales teams’ planning processes. 
  • Industry Marketing Provider: This client was experimenting with pay-per-click and display ads in an effort to develop leads through their website. UA and Google Ads were able to assess ad performance related to site visits and form submissions, but not tie those through to actual leads much less converted accounts. We worked with their CRM to output lead data to be integrated with UA data, and their sales team to report on actual business converted which led to a report with proper cost-per-lead-converted that could also be compared with other business development investments. An added benefit was that this led to an interdepartmental collaboration that previously did not exist.
  • Professional Association: As COVID drove its annual conference remote, this client sought to understand participant engagement throughout their multi-day online event. While UA was able to provide a general sense of interest across sessions, it did not offer the specificity of information the client wanted. By combining UA data with user analytics from the website CMS, video engagement data from the live streaming vendor and social listening data we were able to provide reports that informed both in-conference adjustments and future event strategies.

First Steps to More Sophisticated Web Analytics

So how do we get started? If you are like many others wrestling with GA4 right now, consider taking a step back from the implementation details for a moment to reconsider what you want to achieve with analytics in general. What are your organizational goals, and do your traditional KPIs truly reflect those? Then take stock of the relevant systems and datasets – not simply what has been traditionally tracked through a Google UA method, but what is available today across your systems. 

In our experience, having a guided conversation about the above items will lead to some discoveries and opportunities. In some cases, the next step will be integrating a new dataset into your data analysis tool. For others it may be identifying the tools to use for that integration. But for many the opportunity will be even broader – to assemble a more comprehensive team of partners to manage and leverage the analytics process, and tackle the challenges inherent. 

As noted at the outset, GA4’s new data model and reporting methods work quite differently than the old UA system which presents challenges to existing users. Whether GA4 alone or in combination with other tools provides an effective analytics solution remains to be seen. But ironically, looking beyond the single source of truth we’ve come to depend on in UA may not only give us a better understanding of our users and their behavior, but also may move us in a direction that actually leverages some of GA4s new capabilities to track across platforms.