What Northern Ecommerce Really Thinks About AI

There's a version of the AI-in-ecommerce conversation that happens at large trade shows: well-rehearsed case studies, vendor-sponsored stages, the careful optimism of people who have something to sell. And then there's the version that happens when you get a room full of practitioners and take the microphone away from the exhibitors.

Manchester Digital's third annual Ecommerce Conference on 21 March, at No.1 Circle Square in Manchester, was mostly the second kind. A full day (10am to 5pm) drawing business leaders, technologists, and ecommerce professionals from across the North West. No conference is entirely free of commercial interest, but the ratio here was noticeably in favour of practitioners.

Which is why the read I got from the room matters.

The Honest Temperature Check

The opening sessions set the tone. The agenda covered AI-powered personalisation, voice commerce, headless architecture, omnichannel: familiar enough territory. But the honest note running underneath was that most of the room was still working out what to actually do with these things. Not reporting live deployments with verified ROI. Not in the business of polished before-and-after slides.

That's not a criticism. It's probably where most ecommerce businesses of the kind that attend a Manchester Digital event genuinely are: mid-market retailers, agencies, digital consultancies, Northern independents who've been building their digital operations for a decade. Not Marks & Spencer's data science team. Not Amazon. The people for whom AI is genuinely interesting and potentially significant, but who have limited time and resource to separate the useful from the hype.

The session that generated most discussion was the closing panel: "Automation, AI and the Future of Shopping," with Miya Knight (Director and Publisher at Retail Technology Magazine), Will Clayton (Principal Consultant at AND Digital), Dan Martin (Client Partner at Microsoft), Richard Davis (Principal Product Lead at Auto Trader UK), and Simon Deplitch (E-Commerce Director at CAVU). The consensus that landed in the room was that predictive AI, the kind running quietly in pricing and inventory systems for years, is already embedded in ecommerce in ways most practitioners don't necessarily label "AI", because it just works. The generative AI conversation is different and more open.

The Authenticity Question

The observation I found most interesting came across several different sessions: the concern that AI-generated content, AI-generated personalisation, and AI-mediated customer journeys all risk eroding the authenticity that makes genuine customer relationships work.

This wasn't a Luddite position. The people raising it weren't suggesting AI shouldn't be used. They were pointing at something real: that the ecommerce brands which have built loyal customer bases in the North West have often done it through a distinctive voice, a genuine service culture, and a relationship with their customers that feels like it came from humans who care about the outcome. AI tools that produce output which is correct but generic — the blandification problem — are a threat to that specific kind of competitive advantage.

Mark Leach's session, titled "Integrating AI capabilities with the essential human layer," returned to this theme throughout. The general argument in the room: there's a version of AI-augmented ecommerce that keeps human judgement at the centre and uses AI to scale the things that don't require that judgement, and a version that tries to replace the human judgement entirely. The sessions returned repeatedly to which version practitioners actually wanted to build.

The Northern Perspective

One thing I find consistently valuable about Manchester-based tech events is the insulation from a particular kind of London hype cycle. Not because the North West isn't paying attention to what's happening in global technology (it clearly is), but because the business problems a Manchester retailer is solving are often not the same as the ones being solved by a London enterprise technology team with a significantly larger budget.

The AI tools actually useful to a mid-market Northern ecommerce business in early 2024 are probably not the ones generating the most press coverage. They're more likely to be the automation that saves someone three hours a week on product data entry, the pricing tool that removes the need for a spreadsheet, or the customer service deflection that lets a small team handle the volume of a larger one.

That kind of AI is real, available, and quietly productive. The conference was a useful reminder that most of the industry is living in that space — not in the frontier applications that dominate the write-ups from larger shows. If you want a sense of the gap between the hype and the practice, a practitioner room in Manchester is a reasonable place to calibrate.

For related coverage, see Manchester Tech Festival 2024 and Ecommerce Expo London 2024.