Full article
Overview
Campaign dashboards can look healthy while redemption economics quietly drift off course. That is the awkward bit. If volume is rising but verified purchase quality is unclear, teams can mistake activity for performance and discover the problem only when finance starts asking sharper questions.
This strategy briefing looks at how one UK FMCG team changed what it measured after a promotion appeared to outperform in late 2025, yet spend moved faster than expected. The practical lesson is straightforward: when the proof of purchase workflow is too light, campaign analytics can overstate success. Better integrity measures do not kill growth; they help you see which growth is real.
The situation: a successful campaign with a hidden cost
In late 2025, a UK food and beverage company launched a digital cashback promotion for a new premium yoghurt range. Early reporting looked strong. Redemption rates were more than 30% above forecast within the first two weeks, and the marketing read-out suggested the campaign had found traction quickly.
Then the finance view raised a different question. Budget drawdown was running ahead of the sales uplift expected from retail partners, so the apparent success on the dashboard did not match the commercial picture. To be fair, this is where many teams get caught: standard reporting counted submissions and unique email addresses, but it did not test whether those claims reflected unique, valid purchases. A strategy that cannot survive contact with operations is not strategy, it is branding copy.
The approach: shifting measurement from volume to verification
The team responded by reframing the problem from campaign volume to campaign integrity. Rather than replacing existing marketing analytics, they added a second layer of measurement designed to show whether redemption behaviour matched genuine purchase behaviour. As it stands, that is the more useful operating model for any promotion expected to scale quickly.
The practical change was a redesign of the proof of purchase workflow. Receipt submission moved from a simple upload step to a validation process with clearer checks and reporting. The team began tracking four measures in particular:
They also added receipt parsing through Optical Character Recognition (OCR) so transaction data could be compared automatically at the point of submission. That gave the team a practical option set. They could tighten rules where duplicate behaviour appeared, or keep approval paths lighter where evidence showed low risk. In a strategy call this week, we tested two paths and dropped one after the first hard metric came in. Same principle here: add controls where the evidence justifies them, not because security theatre looks reassuring on a slide.
- Duplicate submission rate: claims linked by matching receipt details, device identifiers or IP patterns, even when different email addresses were used.
- Time to validate: the average time needed to approve a legitimate claim, used to monitor customer friction and operational efficiency together.
- Blocked claim reasons: a coded view of rejections, including duplicate receipt data, suspected image manipulation and claims submitted from outside the UK.
- Suspicious clustering: spikes in claims from a narrow set of devices or network locations, which can indicate organised abuse rather than ordinary customer response.
Outcomes: a clearer view of campaign health
Once the new rules were applied retrospectively to the yoghurt promotion data, the team could compare baseline reporting with validated outcomes. The real picture emerged: an estimated 18% of redemptions were duplicate or fraudulent submissions. Those claims had been counted as campaign success in the original dashboard, even though they represented budget leakage. Growth claims without baseline evidence should be parked until the data catches up, and this was a textbook case.
For the next major campaign in Q1 2026, a digital voucher for a new soft drink, the revised framework was in place from launch. Headline redemption volume came in lower than the yoghurt campaign, which on old reporting might have looked like underperformance. In reality, the new measures showed a healthier result: the duplicate submission rate stayed below 1%, more than 5,000 suspicious claims were blocked in the first month, and campaign spend could be reconciled against verified purchases with far more confidence.
The financial effect was clear. Compared with the projected overspend pattern seen in the yoghurt promotion, the soft drink campaign delivered a direct cost saving of more than £60,000. Just as important, marketing reporting improved in quality. Teams could distinguish between total demand, verified demand and blocked suspicious demand. That made campaign decisions an evidence-led choice rather than a guess dressed up as confidence.
Lessons for other UK promotion teams
The experience of this FMCG brand offers several practical lessons. The first is to treat redemption spikes as a diagnostic signal, not just a victory lap. If volume rises sharply, test duplicate rates and claim velocity before calling the campaign a success.
Next, instrument the proof of purchase workflow from day one. It is far easier to set a baseline at launch than to reconstruct one after a budget issue appears. At a minimum, teams should be able to compare submitted claims, validated claims, and blocked claims by reason.
It is also crucial to design for trade-offs openly. A lighter journey may improve completion rate, but it can also raise the cost of invalid claims. A stricter journey may protect the budget better, but only if it does not introduce enough friction to damage legitimate participation. The right answer depends on the category, incentive size, and likely abuse patterns.
Finally, giving marketing, finance, and operations a shared view of the same evidence is key. When those groups work from separate dashboards, causality gets muddy and decisions slow down. When they can all see the same baseline and outcome measures, the next move becomes much easier to agree.
The next move
If your team is still judging promotional performance mainly by redemption volume and top-line engagement, there is a decent chance part of the commercial picture is missing. The safer option is not to add friction everywhere; it is to measure where leakage appears first, then tighten controls with evidence and timing on your side.
If you want a clearer read on whether your current proof of purchase workflow is protecting margin or merely processing claims faster, Kosmos can help you map the option set and the trade-offs in plain English. We can look at where validation adds value first, what to test next, and how to give marketing and finance one version of the truth before the next campaign goes live.
