Consent mode problems arrive as a phone call that opens the same way: conversions are down thirty percent, nothing changed, and the business says sales are fine. The cause is nearly always a consent change that the marketing team found out about afterwards.
What governs consent mode in the UK
Cookies and similar technologies fall under PECR — the Privacy and Electronic Communications Regulations — alongside UK GDPR. Non-essential cookies, which includes advertising and analytics, require consent before they are set. Not implied consent, not continued browsing.
That means a meaningful share of your UK visitors are, correctly, not being tracked. The question is not how to avoid that. It is how to measure accurately within it.
Consent mode is not a checkbox
Google’s consent mode lets tags adjust their behaviour based on consent state, and when properly configured it allows modelling to recover some of the gap. Meta has an equivalent signal.
The common failure is that the consent platform is installed and the signal is never actually wired to it. Everything looks connected. Nothing is communicating. We find this on a large share of the UK accounts we audit, and it is usually nobody’s fault — the banner was a legal project and the tagging was a marketing one.
A diagnosis in three steps
- Compare platform conversions with your own back end over the same dates. Same orders, fewer reported conversions means measurement.
- Find out when the banner last changed. Ask the web team and the legal team separately; they often have different answers.
- Look at event match quality in Meta, and the conversion diagnostics in Google Ads. These date the change for you.
Why it costs more than the report
Under-reporting is annoying. The expensive part is that smart bidding optimises on the signal it receives. Fewer conversion events means a worse model, which means genuinely worse decisions and genuinely higher costs — not just a worse-looking dashboard.
The reporting loss is cosmetic. The bidding loss is real money.
What to do
Wire consent mode properly and verify it with the platform’s own diagnostic rather than assuming.
Add server-side events where consent permits, with proper deduplication and enough customer data attached for matching to work.
Do not solve it by breaking the law. There is always somebody willing to suggest firing tags before consent. It is a regulatory exposure, and it is not worth the conversion data.
Design the banner thoughtfully. Within the rules there is a large difference between a banner that is clear and one that is hostile. Clear banners get more consent, and they do not create a complaint risk.
The measurement that survives all of this
Hold a blended view at business level: total revenue against total ad spend. Consent cannot touch it, platform double-counting cannot inflate it, and it reconciles with your accounts. Our blended ROAS and MER calculator does it in one screen.
Use platform figures for choosing between ad sets. Use blended figures for deciding budgets. Most of the confusion in this area comes from using the first for the second.
One more habit worth forming: annotate consent and tracking changes in your reporting on the day they happen. Six months later nobody remembers, and that annotation is the difference between a five-minute diagnosis and a fortnight of investigating the wrong thing.
Questions people ask
Why did my UK conversions drop without any change?
Usually a consent banner was installed or reconfigured, and tags now wait for permission that a large share of visitors do not give. Sales are unaffected; what fell is the number of conversions your platforms can see and attribute.
What is consent mode?
A way of telling Google tags whether a visitor consented, so that they either run fully or send limited, cookieless signals that can be modelled. It is a wiring job between your banner and your tags, not a setting you switch on.
How do I diagnose it?
Compare orders in your own back office with reported conversions over the same period, check when the banner was last changed, and inspect what the tags send before and after consent. If back-office revenue is flat, the problem is measurement.
What actually helps?
Server-side tagging and the Conversions API with good match data, consent mode wired properly so modelling can work, and a banner design that earns genuine consent. Beyond that, judge performance on your own revenue rather than platform-reported numbers.
Working with UK brands: Performance marketing for UK brands