Outgrowing Google Analytics: what happens when your questions change faster than your platform

For more than fifteen years, Google Analytics has been the default answer to the question “how do we measure our website?”. It is free, familiar and good at what it was built for: understanding where visitors come from and which campaigns bring them in. For a large share of organisations, that is still exactly what they need, and there is no reason to change.

But analytics platforms rarely fail overnight. More often, the questions an organisation asks change, while the platform stays the same. At some point the gap becomes too wide to bridge with workarounds, extra tools and more training. Recognising that moment early, and knowing what to do about it, is what this article is about.

A team that did everything right

Liantis is a Belgian HR and business services provider for the self-employed and employers. Its customer portal, My Liantis, is not a single website but an ecosystem of applications, from social insurance to payroll communication and workplace prevention, all behind one login. As part of a redesign, the product organisation set itself a clear ambition: build the backlog on evidence of what users actually do and need, rather than on the gut feeling of individual product owners.

They started with the tool they already had. In 2023 we trained their product owners in GA4, and tracking became part of the teams’ definition of done: every new feature had to ship with measurement. On paper, that is textbook data maturity.

Twelve months later, the conclusion was sobering. The product owners were still not getting the answers they needed. GA4’s standard reports are designed around marketing analytics (acquisition, channels, conversions), not around feature adoption, retention or journeys that span several applications. The custom reporting environment proved too cumbersome for non-specialists to build reports they could rely on week after week.

Other gaps added up. Capabilities that mattered for their goals, such as heatmaps and user surveys, lived in a separate tool. Privacy concerns made the team reluctant to send more user-level data into Google Analytics, which in turn limited what could be analysed. And as a B2B portal, Liantis needed to understand behaviour at the level of the client company, not just the individual user.

The data Liantis could collect from the portal with GA4 said too little about who its users were, or how often and in what way they used its features. This was not a configuration or implementation issue. The team had simply outgrown Google Analytics.

Four signs you are reaching that point

Liantis is not an exception. Across the implementations, audits and migrations we work on, the same signals appear long before anyone formally questions the platform.

Recurring debates about data quality. Meetings drift from “what does this tell us?” to “is this number even right?”, often because GA reports something different from internal systems. Once trust erodes, the data stops being used for decisions.

Resistance from privacy and legal teams. When the people responsible for compliance are uneasy about consent, data residency or sending user-level data to Google, analysts end up measuring less than they need to.

Frustration among practitioners. Getting an answer involves manual exports, spreadsheets and doubt about whether a report was built correctly. People stop looking, because they cannot find what they are looking for.

A shift in what the data is expected to do. Digital Analytics moves from a reporting tool to a source for enriching customer profiles, feeding a data platform, or powering personalisation and AI. At Liantis, the data team was building a modern data platform and a recommendation engine. Without consistent interaction data flowing into it, much of that potential would remain unused.

If two or more of these sound familiar, it is worth asking whether your platform still fits, not merely whether it still works.

Start with requirements, not a shortlist

The instinctive next step is to look at alternatives and book demos. We would argue that is the wrong place to start. A process that begins with “which tool should replace GA?” tends to be decided by the most convincing demo or the strongest sales relationship, not by what the organisation actually needs.

At Liantis, the selection began with a workshop to sharpen the requirements the team had already drafted, with input from both product owners and the data team. Those requirements became a weighted scorecard, and only then did we build a shortlist of three vendors to test against it. Product owners, the people who would use the platform daily, joined the demos, because usability for them was a requirement in its own right. And before any contract was signed, the preferred platform ran a proof of concept on Liantis’s own environment, with its own data.

Two factors weighed heaviest in the final decision. The first was having all the required capabilities in a single platform: event analytics across applications, journey and funnel analysis, company-level reporting, heatmaps and session replay, rather than stitching several tools together. The second was the vendor’s vision, in particular its roadmap for AI. For a multi-year commitment, where a platform is heading matters as much as where it stands today.

That combination will look different for every organisation. A regulated enterprise will weigh privacy and data residency more heavily; a large retailer might put warehouse integration and cost at scale first. The point is that the weighting should be decided before you start talking to vendors.

The migration is not the hard part

Organisations often postpone a platform switch because the migration looks daunting. In practice, the technical side proved more manageable than expected for Liantis.

Both platforms ran in parallel throughout the implementation. Because Liantis used the free version of GA4, this came at no extra licence cost and removed the pressure of a hard cutover date. The rollout was phased, product team by product team, rather than a single big-bang go-live.

The loss of historical data, often the biggest objection, mattered less than feared. In a product analytics context, teams focus on the current experience rather than year-on-year benchmarks.

The real challenge is adoption. A new platform only creates value when people change how they work: asking different questions, trusting the answers and acting on them. That takes deliberate training, time for teams to explore on their own, and follow-up sessions once the first real questions surface.

When it works, the effect can be quick. Within the first month after delivery, Liantis’s product owners had already added two new features to the roadmap based on insights from the new platform. That is exactly the evidence-based backlog they had set out to build.

The question behind the question

The lesson from Liantis is not that Google Analytics is a bad tool. It is that “which analytics platform should we use?” is really a question about what your organisation needs its data to do. When that answer changes, from reporting on campaigns to understanding product usage, from dashboards to data that feeds other systems, the platform deserves a fresh evaluation. That includes any successor your current vendor is steering you towards.

The organisations that come out ahead will be the ones that get specific about their requirements before they get attached to a shortlist.

To help with that, we have turned the requirements we see most often into a practical framework: 24 checkable criteria across six categories (data ownership, privacy and compliance, warehouse integration, AI readiness, cost transparency and usability), plus a step-by-step evaluation process.

Download the white paper Does your analytics platform still fit?

Stichd is sponsor of DDMA Digital Analytics Summit 2026. Digital Analytics Summit took place on 1 October in Hotel Casa Amsterdam.

Jente De Ridder

Managing Partner and Founder | Stitchd

• LinkedIn

More about Jente

Jente De Ridder is Managing Partner and Founder of Stitchd, a data and analytics consultancy that helps organisations across Belgium and the Netherlands design, implement, audit and migrate their analytics stacks, platform-agnostic by design.