At eCommerce Expo 2024, Nobody Was Pretending

One of my favourite conference exercises is calibrating how far the room actually is from the thing being discussed. At NRF, the gap between keynote vision and production reality tends to be measured in years. At smaller practitioner events, it's sometimes weeks or months. At eCommerce Expo London, which occupies an interesting middle ground between trade show and conference, the gap is usually measured in honest uncertainty.

The 2024 edition, held at ExCeL on 18 September, was dominated by AI. But the dominant note in the room wasn't confidence. It was something closer to honest bewilderment. Which, at this particular moment in the AI commerce story, feels like the most accurate possible response.

The Revolution Announcement

Benjamin Lang, Head of Northern Europe, Partnerships at Shopify, opened proceedings with a line that landed in the room the way these things land: "A revolution is coming and it's called AI." Some people nodded along. Some people wondered what specifically they were supposed to do about it by Monday morning. The sentiment is hard to disagree with. The operational specificity is harder to extract.

What was notable, across the sessions I attended, was the retailers' own willingness to admit that the AI experiment is very much ongoing. The mood the RTIH review captured was one of reluctance to claim progress, with practitioners noting that the technology was moving faster than their evaluation cycles could keep pace with. That's a more honest summary than most conference retrospectives produce. The speed of development is generating a specific kind of anxiety: not "we don't think this is important" but "we can't assess what to invest in because the landscape is shifting faster than our procurement timelines."

Lang's own follow-up on stage made the structural point clearly: AI without data integration is just noise. "A unified commerce approach that spans all of a retailer's channels is still needed," he said. "AI that connects to all channels and into all systems will be much more powerful than that without. Otherwise, it's a case of garbage in, garbage out if data mastery and integration aren't properly established first."

That framing is right, and it's frustrating that it still needs saying in 2024. The vendor pitch is almost always the AI capability. The real question is whether the underlying data architecture is in good enough shape to make the AI useful. At most mid-market UK retailers, it isn't yet.

Post-Purchase and the Four Cs

The session I found most substantive was in the Post-Purchase: Optimising the Delivery Experience stream, where Stacey van den Aardweg, Director of UK Fulfilment at Amazon, worked through a framework she called the four Cs: Consumers, Cost, Complementary technology, and Cyber security.

It's not a glamorous framework. Post-purchase rarely is. But it maps well onto where AI is currently delivering genuine, measurable value for UK retailers. Customer service automation, returns processing, delivery exception handling, dispute resolution: these are high-volume, structurally repetitive interactions. AI performs well on them. And they're also the interactions with the most direct bearing on repeat purchase and net promoter score.

There's an argument, which I find fairly compelling, that AI-powered post-purchase experience is a more immediately accessible first deployment for mid-market UK retailers than AI-powered discovery or personalisation. The data requirements are more tractable. You don't need a sophisticated customer data platform; you need clean product data, clean order data, and a competent customer service automation layer. The integration work is bounded in a way that AI-driven personalisation at scale is not.

Van den Aardweg's point on the Consumer element of the four Cs was direct: "What matters is offering choice." It sounds basic. Applied to AI-powered post-purchase tooling, it means the automation shouldn't route every query to a bot and call that service. Choice, in this context, means intelligent triage that offers human escalation where the query warrants it.

The Vendor vs. Practitioner Dynamic

eCommerce Expo is a large commercial event and it has the vendor presence you'd expect. The AI content that was most useful was from actual retailer teams talking about what they're running, why it is or isn't working, and what they've learned.

The less useful content had a predictable shape: vendor-sponsored sessions where the framing is the platform's capabilities rather than the business problem, and case studies selected for being exceptional rather than representative. I don't mean this as a particular criticism of eCommerce Expo; it's a structural feature of commercial event sponsorship. But it does mean you have to filter actively, and the filtering requires knowing enough to tell the difference between a genuine use case and a vendor story.

The genuine peer learning at events like this tends to happen in corridors, between people who've tried the same tools and have views that aren't shaped by a non-disclosure agreement. That's worth making time for, even though it doesn't appear in the session recordings.

The 2024 Temperature

My read from eCommerce Expo 2024: the UK ecommerce community believes AI is significant and is investing in understanding it. It is not yet confident about what to do with it. The pace of change is creating genuine uncertainty about which bets to make. And there is a reasonable cohort of retailers who have been through enough hype cycles to calibrate their enthusiasm carefully, including voice commerce, AR shopping, and conversational chatbot technologies that promised transformation and delivered modest improvements at best.

That sceptical-but-engaged position is the right one for mid-market UK retailers in late 2024. Not "ignore AI" and not "bet everything on the most-talked-about application." Something more like: build the foundations, run bounded experiments, don't freeze waiting for certainty, but don't accept every vendor case study as representative of what you should expect.

Which is, it turns out, genuinely hard to hold. It's much easier to be enthusiastic or dismissive than to sit in productive uncertainty. That a room full of UK retail practitioners was willing to do it openly, on record, is about the most useful thing eCommerce Expo 2024 produced.

---

Further reading: the RTIH review of eCommerce Expo 2024 covers the Post-Purchase stream in detail.