Today’s shoppers don’t move neatly from awareness to consideration to conversion. They jump between TikTok videos, Google Shopping carousels, AI-generated answers, marketplaces, and brand sites, often in the same five-minute window. Discovery and conversion happen everywhere at once, and the retail funnel that ecommerce teams have built around for two decades no longer holds.

That was the premise of “The New Retail Funnel,” a panel webinar hosted in May 2026 by The Fashion Network, Ecommerce Club, and Athos Commerce, the intelligent discovery platform for ecommerce brands.¹ Five practitioners spent an hour pulling apart what the path to purchase looks like now: Stephanie Brown of Athos Commerce, Rosie Bentham of TikTok Shop UK, ecommerce consultant Paul McDermott, fractional merchandising consultant Margaret Simm, and fractional Head of Ecommerce Alex Green.

Based on the discussion, it became clear that retailers no longer compete on stock alone. They compete on visibility across every channel where shoppers make decisions, and three changes are driving that competition. Each one changes what your team should be doing about it.

Discovery is now everywhere at once

Discovery used to start with a brand site, a search result, or a paid social impression, all of which sent shoppers to a measurable destination. Nowadays, it starts with an AI assistant, a TikTok video, or a generative answer embedded in a Google search. The shopper learns enough to make a decision before they ever touch a website where you could track them, and the product that ends up in the cart is usually one that the AI agent or the creator recommended on the way.

Paul McDermott described the change in product content terms. Good SEO practices still apply, but keywords and bullet points alone are no longer enough. Product information has to be conversational, carry more context, and live in more places than the product detail page. FAQ pages, help content, and customer service transcripts all serve as source material for the large language models (LLMs) that AI assistants draw on. What you can’t control is how your product appears on Reddit or in a creator’s TikTok review, so consistency across all owned channels matters more than it used to.

Stephanie Brown framed the same change from the product data side. The consumers of your product content are evolving. You are preparing product data for what amounts to a digital sales force. AI agents reason over your product information and make purchase decisions on behalf of customers with very different intent profiles.

Stephanie’s kitchen-appliance example illustrates the point. The questions a shopper asks ChatGPT when buying a kettle to match a kitchen finish differ from those a 76-year-old buyer would ask, and both differ from those a renter buying a first kettle would ask. The product data has to contain enough metadata to answer all of them. Brands that wrote feeds for keyword matching have to rewrite them for conversational reasoning, or they don’t show up at all. Eighty-three percent of ChatGPT’s product carousel already comes from Google Shopping data, so feeds built for traditional search now also feed AI assistants.

Creators became the new merchandisers

The AI side of the funnel reshapes how shoppers ask questions, and the creator side reshapes how they decide. Rosie Bentham, who leads Strategic Key Accounts for Fashion and Beauty at TikTok Shop UK, called TikTok a content-discovery-led platform. That matters because the content that drives sales on TikTok is structurally different from the content brands have been making for years.

The content that converts on TikTok is, in Rosie’s words, authentic. Users see a creator using a product, hear them solve a problem with it, and watch them push a clear call to action while the video plays. The “aha” moment lands in the first three seconds. Aspirational visuals that work on Instagram often underperform on TikTok because users want honest commentary from real people demonstrating the product.

The structural change is that the authority has moved. The creator with 50,000 engaged followers is the new merchandiser, and the livestream host doing two hours of product demos is the new sales floor. For TikTok Shop key accounts, affiliates and creators drive 60 to 75 percent of performance, Rosie said. (For a related view, see The TikTok Shop margin trap.) The investment question has expanded to include how much to spend on creator partnerships and content volume.

Stephanie added the platform side. Athos recently went live with Argos and John Lewis on TikTok Shop, two UK retailers piloting the channel to reach customers their existing storefronts couldn’t capture. Content quality is one input, but the catalog has to be in sync, the inventory has to update in near real time, and the pricing has to match. A brilliant creator video for a product that has gone out of stock between the seller’s system and the TikTok feed is the most expensive kind of friction.

The dashboards that no longer match the funnel

AI discovery and creator-led commerce both create measurement problems, as Paul McDermott put it plainly. Both happen off your owned channels, in spaces you can’t fully tag. Attribution is getting harder, and detailed last-click attribution is not going to recover.

Paul argued that nobody becomes more profitable from a great-looking dashboard. The decisions teams make matter more than the report layout. The two metrics he flagged as overlooked were customers and net cash. Knowing which products bring in new customers, which products repeat customers buy, and how buying patterns evolve matters more than chasing the granularity cookies and Facebook pixels can no longer deliver.

Alex Green pushed the same argument from the operator’s chair. Rolled-up metrics like average order value, total visitors, and blended conversion rate flatten exactly the signal a trading team needs. Average order value may have climbed because you sold more, or because you sold fewer high-volume low-price items. Conversion rate may have improved because of better user experience, or because you turned off a discovery channel that was bringing in early-funnel visitors and watched the rate climb without underlying changes. Trading teams need to conduct analysis by channel, product, and customer segment to find the signal.

In practice, ecommerce leaders need to invest in customer-level reporting and stop chasing pixel-perfect attribution. Google’s Merchant API now blends impressions from traditional Search, Gemini Search, and AI Mode, so even Google’s reporting is becoming less granular by source.² A trading-period mindset works better than rebuilding a single source of truth that AI has already moved past.

When merchandising, marketing, and ecommerce have to talk daily

Margaret Simm has spent years inside trading meetings at fashion and apparel retailers, and those meetings have changed. They used to focus on online and physical store channels. Now, the same meetings span product, marketing, customer service, social, and tech. Trustpilot and review trends sit in the same agenda as buy plans, and AI usage gets discussed on Monday morning alongside hero product performance.

Paul reinforced the same point from the larger-organization vantage. Functional verticals still exist, each with its own heads of ecommerce, marketing, and merchandising. The retailers who win are the ones whose teams cross those lines casually, using shared dashboards, planning cadences, and incentives, rather than treating AI or social commerce as another walled-off team.

Alex Green told a story that captured the cost. A retailer had a slow-moving product sitting in inventory for years, with the owner saying he would be dead before they shifted it. The team cleared it. Then, because the institutional memory of why it didn’t exist in any shared system, the buying team reordered it.

Margaret made the same point from a different angle. Customers who shop on TikTok may behave differently from customers on the website, so promotional planning, stock allocation, and creator briefs have to talk to one another. To succeed in Discovery Commerce, Rosie added, you need buying, product, ecommerce, social, influencer, advertising, and media teams in the same conversation. Siloed teams produce underperforming channels even when the product is strong.

From managing stock to managing discoverability

Margaret Simm distilled the panel’s argument into a single line.

That changes the work of every team. Merchandisers allocate visibility as well as inventory, and marketers shape the inputs AI agents and creators rely on, alongside their traffic work. Ecommerce leaders own the unified data layer that keeps stock, pricing, content, and reviews accurate across every channel where decisions get made.

Stephanie’s closing summary tied the change back to the asset most teams under-invest in. Product data is a strategic commercial asset, and the retailers who treat it that way pull ahead of the rest.

Paul offered the final grounding note. The acronym list keeps growing, from agentic commerce and answer engine optimization (AEO) to generative engine optimization (GEO) and Discovery Commerce. Underneath them, what shoppers want hasn’t changed. Knowing the customer, earning trust, and showing relevance worked at a market stall, and they still work when an AI agent runs the search.

Where to start this quarter

1. Audit your product data for AI readiness: Run a feed audit against the conversational questions shoppers in your category ask today. The keyword set you optimized for two years ago no longer covers the discovery pattern.
2. Pick one off-site channel to seriously invest in: Most teams are spread thin across five social platforms. Focusing spend and creator partnerships on the one that best fits your customer demographic yields a stronger signal than a thin presence everywhere.
3. Re-segment your reporting around customers, not just channels: Move at least one weekly meeting away from rolled-up dashboards and toward new-versus-repeat customer behavior across the channels that matter most.

The teams that thrive in this new map of discoverability will treat product data, creator partnerships, customer segments, and cross-functional planning as a single, connected system. Athos Commerce calls that intelligent discovery. The panel called it the new retail funnel. Both descriptions point in the same direction.

Sources & Further Reading

1. “The New Retail Funnel: How to Win Product Visibility in the Age of AI.” Panel webinar hosted by The Fashion Network, Ecommerce Club, and Athos Commerce, May 20, 2026.
2. Google. “Merchant API.” Google for Developers, accessed May 2026. https://developers.google.com/shopping-content/guides/quickstart.