RetailX 2026 Report: Reliable forecasting from mixed signalsEstimated reading time: 2 minutes
Whitepaper

RetailX 2026 Report: Reliable forecasting from mixed signals

By: Epsilon Marketing | July 13, 2026
Title slide for a RetailX presentation: "Reliable forecasting from mixed signals", with subtitle "Data workbench: The new rules around finding customers and building loyalty". Associated with Epsilon, dated 2026.

Marketers have more customer data than ever, yet fewer reliable ways to connect it into a coherent view of the shopper. Discovery has fragmented across search, social, retail media, CTV and AI assistants, and the response to each new channel too often makes the underlying problem worse.

This report, in partnership with Internet Retailing, looks at what UK consumers are actually doing, where recognition breaks down, and what that means for the cost of finding and keeping the right customers.

Key findings

  • 68% say the brands they buy from regularly already know what they want from them. But 53% are bothered when a brand treats them as a new customer despite previous purchases, and more than 65% limit or reject cookies. Recognition is expected, but the digital signals that enabled it are being withdrawn.
  • Almost 80% of consumers say repeated ads become annoying, and 73% say excessive repetition damages their view of a brand. Over-frequency looks like a media problem but is more accurately an identity one.
  • 71% agree that being logged in or a loyalty member results in more relevant communication, emphasising the value of person-based data and the tangible value exchange customers experience.
  • Perceived product quality climbed from 38% to 47% between January and May as an everyday purchase influence, while low price fell from 57% to 54%. Brand work still earns its keep, even under pressure on discretionary spend.
  • 74% would not want AI tools making purchase decisions on their behalf, and more than 70% still verify AI recommendations by visiting retailer sites or reading reviews before buying.

What you'll learn

  • Why acquisition costs keep climbing even as data volumes grow.
  • What CTV solves, and what it complicates.
  • Why the useful question is no longer which channel converted, but which combination did.
  • How over-frequency exposes weak identity, not weak media planning.
  • Why value exchange sits underneath the strongest identity frameworks, and where loyalty programmes fit.
  • How retention data is quietly becoming the sharpest acquisition tool marketers have.
  • What happens to a category when consumers let AI assistants choose the retailer.

How brands are responding

The report features work from Currys, Cheerz and Fat Face, showing how first-party data, identity and coordinated activity across channels are helping brands turn fragmented signals into measurable customer growth.

Download the report

For the full survey data, alongside Epsilon's perspective on why rising acquisition costs are really an identity problem.