Personalisation & AIPart 4 of 82 min read

AI personalisation in retail, beyond the trend lists

Every 2026 trends article says AI personalisation is now a baseline expectation. Few explain what actually changes for the people running it.

Rob GeerdinkRob GeerdinkFounder and CEO, Personalisation Hub
3 November 2026
Store associate using a tablet in a supermarket aisle

In short

AI changes how personalised campaigns are made, not just how they’re targeted.

  • Agents draft and launch; marketers edit and approve.
  • Live context replaces fixed segments.
  • Closed-loop learning and data ownership matter most.

If you’ve read one “AI trends for retail” list this year, you’ve read them all. AI personalisation, tick. Chatbots, tick. Predictive inventory, tick. The common thread is that personalisation has moved from competitive advantage to baseline expectation.

Fair enough. But “use AI for personalisation” isn’t a strategy. Here’s what I think actually changes.

Authoring becomes a conversation

Today, a personalised campaign means briefs, segments, creative for every channel and weeks of back and forth. In the AI-native model, a marketer describes what they want to an agent. The agent looks at past performance, suggests the experience, builds the creative for every touchpoint and launches it. The marketer edits and approves instead of producing.

Segments give way to context

Segments were a workaround for not knowing enough, quickly enough. When a system can see the customer’s status, the store, the stock and the moment, it doesn’t need to put them in a bucket. It decides one interaction at a time.

Segments were a workaround for not knowing enough, quickly enough.

The loop closes on its own

The real prize isn’t smarter targeting. It’s a closed loop, where every experience, interaction and sale feeds back into the system so the next decision is better than the last. That’s what turns personalisation from a campaign into an operating model.

Data ownership stops being optional

All of this runs on your customers’ data. Running it in someone else’s cloud, on someone else’s terms, is a risk more retailers are unwilling to take, especially now that open models make it practical to run AI inside your own environment.

Teams change shape

Fewer people producing creative variants. More people setting guardrails, reviewing what the agents do and deciding what good looks like. That’s a skills shift, and it’s the part most retailers underestimate.

Related on Personalisation Hub

You’ve read part 4 of 8

Up next

When agents do the shopping, what is the store for?Read part 5 of 8

See what personalisation looks like in your stores.

Arrange a Demo

Enterprise Personalisation Platform