29/09/2026
At UCA26, eight partners each hosted a round table in ten minute rotations and spent the afternoon listening. Here is what they heard, and what to do about it.
1. AI is in production, but the room is split in two. There is a leading edge with live use cases, a long tail that has not started, and very little in between.
What to do: move AI out of content and into operations, such as diagnostics, data quality and internal knowledge, where the return is measurable.
2. Product data decides discovery. Shoppers are finding products on marketplaces, shopping surfaces and AI assistants, and if your catalogue is weak your brand does not show up at all.
What to do: fix the product data before buying more channels. Give the catalogue an owner and a quality standard, and start tracking sales from AI assistants.
3. The blockers are internal. Marketing and service work from two different views of the customer, big brands are waiting on executive buy-in, and some retailers still run fulfilment on spreadsheets.
What to do: get service and marketing onto one view of the customer, turn on AI chat for the hours you do not staff, and audit fulfilment before peak.
4. Growth has to come out of margin. The international demand is there. Shipping, duties, returns and FX decide whether it pays.
What to do: validate a market on the traffic and orders you already have before spending on it, and add up what you lose to bank transfers, platform fees and FX.
5. Retailers want a shortlist. The request we heard most all day was “just tell me the three things to focus on”.
What to do: ask ChatGPT, Claude and Perplexity ten real buying questions while logged out, and check 200 products to get your defect rate.
The bright spot was that plenty of retailers said sales were up, when most had expected gloom.
