How to set KPIs for a digital experience that is not a conversion funnel

Most KPI frameworks are built around one thing: a purchase. Reach, click, conversion, sale. That works well for e-commerce or a lead generation campaign. But what if you are building something that was never meant to work that way?

A loyalty platform is designed to encourage behaviour, not force a single transaction. A brand activation is meant to create emotional connection, not drive immediate revenue. A community platform exists to bring people back repeatedly, not push them through a funnel.

When you apply standard funnel metrics to these kinds of experiences, you measure the wrong things. And what you measure determines what you build. At Livewall, we see this problem come up repeatedly with ambitious brands: the experience is built with the right intention, but judged against the wrong measure.

Step 1: start with the behaviour you want to change

Ask yourself: what specific behaviour does this experience need to encourage? Not 'more engagement' or 'stronger brand affinity'. Something concrete. Daily return visits. Discovering multiple product categories. Referring a friend. Using a loyalty card more often.

For the HEMA Stapelgek loyalty activation, the target behaviour was clear: daily app interaction tied to physical store visits. KPIs were therefore return frequency, weekly active users, and the percentage of players who made a purchase after participating. Not click-through rate on a banner.

That distinction sounds simple, but it changes everything: the brief, the product decisions, the reward structure, and the way the team experiences success.

Step 2: measure engagement quality, not just volume

Pageviews and session duration are lazy KPIs. They tell you people showed up, not that they got anything out of it. For non-funnel experiences, you want to know whether engagement is deep and repeating.

Better indicators:

• Return rate — are people coming back without being pushed?
• Interaction depth — do they reach multiple layers, or drop off after the first screen?
• Social spread — do they share the experience unprompted?
• Micro-action completion — do they finish challenges, collect badges, complete their profile?

For the Decathlon always-on loyalty campaign, repeated app opens and the number of linked movement sessions were better measures than classic conversion figures. Members who returned weekly had a demonstrably higher lifetime value, even when that was not visible in standard campaign reports.

These are the kinds of consumer engagement trends that separate genuinely effective programmes from campaigns that look successful until the next quarter.

Step 3: connect engagement signals to commercial outcomes

This is where the real challenge sits. Measuring engagement is relatively straightforward. Drawing the line to revenue, retention, or brand value requires more thought, but it is essential for defending what you build internally.

A practical approach is segmentation analysis: compare the purchasing behaviour of active participants against non-participants over the same period. This is not perfect attribution, but it gives direction. With Proximus+ World we observed that users who actively engaged with the interactive brand world churned at a significantly lower rate than comparable users who did not. That signal has real value, even without a perfect causal chain.

For community experiences like Sportvisunie, the commercial proxies look different: membership renewal rates, referrals to partner products, and active knowledge contributions are better indicators than page visits.

Step 4: use leading indicators, not only lagging ones

Conversion rate and revenue are lagging indicators: they tell you what happened after the fact. For non-funnel experiences, you also need leading indicators that help you steer earlier in the process.

A solid measurement setup combines both:

Leading (early signals):

• Day-1 retention after first participation
• Onboarding flow completion
• Social sharing actions within the first 48 hours
• Profile or account completeness

Lagging (commercial outcomes):

• Average order value for active participants vs. control group
• Churn rate following campaign participation
• Net Promoter Score movement
• Repeat purchases per quarter

For brand activations like Wehkamp Wanna Have Days, daily return rates are an excellent leading indicator: if people come back without a reminder, the activation is working. The lagging indicator, incremental revenue during the campaign window, follows naturally.

Step 5: make the KPI structure defensible internally

The best KPI framework means nothing if stakeholders do not understand or trust it. Make sure you translate engagement goals into language the business speaks.

'Daily active users grew by 40%' is less convincing than 'Members who returned daily purchased on average 2.3 times more during the same period than inactive members.'

This requires extra preparation up front: you need to set up your measurement plan before you build, not after you launch. Segment your users, establish control groups where possible, and document which behaviours you intend to influence. That sounds like more work, but it is the only way to demonstrate what the experience actually delivered.

At Livewall, we start every project with a digital strategy session where we do exactly this: define KPIs before the first concept exists. What behaviour do we want to change? How do we measure it? How do we connect it to commercial value? That conversation upfront prevents a lot of argument later.

What this means in practice

Non-funnel experiences are not harder to measure than funnel experiences. They just require a different starting point. Begin with behaviour, not channel activity. Measure engagement quality, not just volume. Connect signals to commercial outcomes through segmentation. Use leading indicators to course-correct, and lagging indicators to report.

The consumer engagement trends we see across the brands we work with all point in the same direction: brands that measure what people do rather than what they click build better experiences, and can prove it.