UK advertisers face a measurement gap as AI and mobile spend climb
According to TechRadar, UK marketers need clearer mobile-grade measurement to prove AI-driven ad spending pays off, as IAB UK forecasts £45 billion in digital spend.
TechRadar reports that UK marketers are entering a pivotal period. Digital advertising budgets keep growing, retail media and connected TV are taking up more space in media plans, and artificial intelligence is moving out of pilot projects and into routine campaign work.
According to the outlet, the central question has not changed: can marketers show that each pound spent produces measurable value? TechRadar argues that this question is now more pressing than ever, as spending grows and automated tools take on more of the decision-making.
TechRadar cites IAB UK's forecast that digital ad spending in the UK will reach £45 billion this year. The outlet notes that mobile accounts for more than two thirds of that spending and that online retail media is among the fastest-growing channels.
The article also says research points to measurement fragmentation as a major issue for the industry. It does not describe the specifics of that fragmentation in detail, so the scale and causes of the problem beyond that point are not clear from the reporting.
Taken together, TechRadar's argument is that AI-driven campaign tools will struggle to demonstrate return on investment unless the underlying measurement is consistent, particularly across mobile devices.
Background
Digital advertising measurement has long been difficult. Marketers want to know which ads led to a sale, a sign-up or a visit, but the path a customer takes often crosses several devices, apps, websites and platforms. Each platform typically reports its own results, and those reports do not always agree. This gap between platform claims and independent verification is often described as fragmentation.
Mobile has become the main environment for consumer attention, which makes it central to any measurement strategy. Phones run apps that collect different signals than web browsers, and operating system privacy rules have changed what identifiers advertisers can use. As a result, many marketers have shifted toward aggregated reporting, modeled conversions and controlled experiments.
Retail media has grown as retailers sell advertising placements on their own websites and apps, using their customers' purchase data. This can offer closer links between ads and sales, but it also means a retailer is often both the seller of the ad space and the source of the performance report. Connected television adds another layer, combining video formats with audience data that can differ from one provider to another.
Artificial intelligence now plays a role in bidding, audience selection and creative testing. These systems optimize toward the signals they receive. If those signals are incomplete or inconsistent, the optimization can be efficient in the wrong direction. Marketers often rely on incrementality tests, which compare exposed and unexposed groups, to check whether a campaign truly adds to sales rather than capturing purchases that would have happened anyway.
Why it matters
Advertising budgets are a major cost for many UK businesses, and finance leaders increasingly expect clear evidence of returns. If measurement is fragmented, the question of which channel deserves more money becomes harder to answer with confidence. Decisions may then favor whichever platform reports the most favorable numbers, rather than the one that performs best.
The issue is especially relevant as AI tools become standard. Automated systems can reallocate spending quickly, but they cannot correct for poor data on their own. A flawed measurement setup can lead to confident-looking dashboards that hide weak results, and that risk falls hardest on smaller brands and agencies that lack the resources to audit platform claims.
Measurement also touches consumer privacy. Better tracking can mean more personal data is collected and matched across devices, while privacy rules and user expectations push in the other direction. Building measurement that is both accurate and respectful of those limits is a practical challenge, not just a technical one. The outcome will shape how much trust the industry can hold with the public.
What to watch
Watch whether the £45 billion forecast from IAB UK holds and how the mix between mobile, retail media and connected television shifts over the coming months. Forecasts of this kind are often revised as market conditions change.
Pay attention to whether industry groups agree on common measurement standards that work across platforms. Shared definitions would make results easier to compare, though agreeing on them has historically been slow.
Retail media is another area to follow, particularly how retailers share data with advertisers and whether independent verification becomes more common. Regulators and privacy authorities may also influence what data can be used for measurement.
Finally, watch whether AI-driven campaign tools produce clearer evidence of return. Marketers who run controlled tests will be well placed to tell the difference between genuine gains and results that only look strong on a dashboard.
Viewpoints
Commentary from our columns: three angles on this story.
The Ledger the business angle
Every marketing budget is ultimately a bet, and the people approving it want to know the odds. When TechRadar describes the pressure to prove that each pound produces value, it is describing a shift already happening in finance meetings across the UK. Chief financial officers are less interested in impressive campaign metrics than in whether revenue moved in step with spending.
Fragmented measurement creates an uncomfortable incentive. If each platform grades its own work, budgets tend to flow toward whoever reports the cleanest story. Retail media and connected television are growing quickly, which means the stakes of that dynamic are rising too. Businesses that insist on a single, independent view of performance may find they can spend less and learn more.
AI adds speed to this equation but not necessarily judgment. Automated buying can move money across channels in hours, which is valuable only if the signal guiding it reflects real returns. The companies that benefit most may be those that treat measurement as a core business function, with the same discipline applied to a finance ledger, rather than a reporting detail handled after the campaign ends.
The Builder the engineer's view
From an engineering standpoint, measurement is a data pipeline problem before it is a marketing problem. Events have to be captured reliably on mobile apps and websites, matched to campaigns without leaking data, and attributed in a way that survives changes to device identifiers and browser rules. Every break in that chain becomes a blind spot.
AI optimizers are only as good as the signals they receive. If conversion events arrive late, duplicate, or with inconsistent definitions across platforms, the model learns the wrong lesson with great confidence. Engineers working on these systems should treat data quality checks, reconciliation between platform reports and first-party records, and clear event naming as basic requirements.
The most useful habit may be the least glamorous one: build holdout groups and run controlled experiments regularly. A well-designed test tells you whether a channel adds value, which no dashboard can fully promise. Teams that combine clean data, honest tests and a healthy skepticism of automated recommendations will be in a far better position than those chasing the newest model.
The Skeptic open questions and weak points
The argument in TechRadar's analysis is persuasive, but it is worth asking what it leaves out. The piece presents measurement fragmentation as a major problem without describing its scale or causes in detail. Readers should ask what the underlying research actually found, who conducted it, and how it defines a measurement failure.
The forecast of £45 billion in digital spending deserves the same scrutiny. Forecasts reflect assumptions about the economy, consumer behavior and platform pricing, and they are often revised. Mobile's share of that spending is significant, but the article does not show how much of it is currently measured in a way marketers trust.
Then there is the phrase mobile-grade measurement itself. It sounds precise, yet the industry has not settled on one meaning. Before accepting that AI cannot deliver on return on investment without it, the open questions are simple: how much improvement would better measurement actually produce, and would the answer be the same for every business? Those questions are worth asking before any new tool is bought.