The ecommerce discovery reset: How AI is changing the rules of selling
- Written by Matt Warren
AI is the first technology in 30 years that I truly believe will make a real difference. I believe AI will genuinely change the game for sellers, consumers, and everyone in between. We’re watching it unfold in real time, and here’s what sellers need to know to stay ahead.
Conversational discovery is the new norm
Shoppers are no longer typing generic product searches. They’re having conversations with AI about exactly what they want. The days of keyword optimisation are giving way to actual conversations about products.
Instead of searching “white t-shirt”, shoppers might ask for “a white t-shirt that will match my denim jeans and won’t be too hot on a summer’s day in New York”.
Then there’s the follow-up. Depending on the user’s history, a chatbot might then say “As you’re a university student, would you prefer a more casual shirt for wearing to class, or something fancier for your evening dance classes?” or simply “Here are some suggestions based on your saved preferences for breathable fabrics with a vintage feel”.
That’s not all. The browsing experience now extends to purchase via chatbots, with Walmart, Etsy, and Shopify-powered sites. Meaning - shoppers will browse and purchase via chatbots, so getting your strategy right is more important than ever.
This isn't just changing how people find products - it's changing what information sellers need to provide. Chatbots run on data – and they need it in real time. Too often, we see sellers use product information that’s 6–12 months old, leading to broken links and outdated listings, which kills the chance of AI discovery.
Actions sellers can take:
Enrich your product intelligence
Go beyond specs: Include context, use cases, questions answered, comparative attributes.
Add alternative language around product trade-offs, “if you want X, consider Y.”
Structure for machine use
Use consistent attribute taxonomies across all channels.
Ensure data fields are machine-readable and versioned (so updates propagate).
Write with the conversation in mind
Design descriptions that anticipate shopper phrasing (“Is this waterproof?”)
Use modular copy blocks that AI can pull into responses.
Visual commerce
The future of product discovery is visual-first. In TikTok and social commerce, short video drives dramatic conversion lift; in fact, up to 70% of GMV on TikTok Shop comes from short video formats.
But visual commerce isn’t just about flashy content — it’s about structured visual data. 3D views, AR previews, multi-angle shots, and synchronized video + spec overlays become the raw materials AI uses to match user intent.
What sellers must do now:
Treat video as part of your standard product launch process.
Use consistent visual taxonomy across stills, 360° spins, and video.
Experiment with AR previews (even in limited categories) to future-proof.
Why operations are part of the new discovery game
AI isn’t only changing how customers find products, it’s changing how they choose who to buy from. Delivery promises, stock accuracy, and fulfilment reliability are now as influential as price or imagery in buyer decision-making. Operational data has become marketing data.
When sellers integrate real-time order, inventory, and carrier data across their systems, that reliability shows up at the point of discovery. It’s not just about speed – it’s also about consistency and visibility.
Actions sellers can take now:
Integrate your shipping and inventory systems: Ensure your ecommerce and fulfilment data update in real time.
Automate dispatch rules: Assign carriers based on cost, weight, and location to protect both delivery times and margins.
Track fulfilment profitability: Fast shipping shouldn’t erode profits – monitor shipping costs against order value daily.
Prioritise predictability: Shoppers value reliability over unrealistic speed. Build systems that deliver on your promises every time.
As fulfilment reliability becomes a visible differentiator, the operationally prepared sellers will be the ones AI-driven systems and shoppers trust most.
Beyond 2025
Looking ahead, early signs point to multimodal and contextual commerce as the next change to shopping habits:
Multimodal search: Combine image + text + voice in one query. AI needs to connect your data dots in real time.
Contextual commerce: AI will adjust recommendations based on location, weather, upcoming events on a user’s calendar.
In short: your data must be flexible enough to respond to context — not just static listings.
What this means for sellers
Think in use cases, not categories
Your product might live in unexpected realms. A screw becomes “the hinge fix for flush cabinets” in conversational contexts. Don’t silo by vertical — imagine cross-context relevancy.
Build modular content systems
Your product stories should be deconstructed: Bullet specs, emotional hooks, scenario descriptions, video, FAQs. That way, AI can assemble them dynamically for different queries.
Be channel-agnostic and data-centric
Algorithms will rise and fall; structured, portable data survives. Build systems that let you move into new channels quickly with minimal rework.
Iterate small, measure fast
Let pilot tests guide adoption: Try conversational listings, visual previews, or AR in small segments before scaling. Use real-time feedback to adjust, not quarterly retrospectives.
The winners in this new era won't be the ones with the biggest AI budgets - they'll be the ones who understand how to make their products discoverable in whatever way people choose to shop.
The bottom line
This won’t all happen tomorrow, but the pace of change is rapid. Sellers who start small, stay informed, and make smart investments will be better prepared than those who wait. Whatever technology lands next year, always:
Research and monitor: Test ChatGPT, Claude, and Perplexity to see how your products show up.
Optimise your data: Keep it detailed and current.
Stay informed: AI capabilities shift fast.
The question isn’t whether to adapt – it’s how to adapt in a way that makes sense for your business.
