Mike McClary stopped selling his Guardian LTE Flashlight around 2017. The heavy-duty outdoor light had been one of his best performers, but he moved on to other products. For years afterward, customers kept emailing him asking where they could still buy it.
That kind of feedback - discovering what customers want based on what they request after you've stopped selling it - used to take years to surface. AI-powered sourcing tools are now trying to identify that signal before the product disappears from shelves, and it's changing how small online sellers decide what to make before committing to inventory.
From Gut Feeling to Market Signals
Small product businesses have always placed bets on what to make or stock. A solo seller running an outdoor gear brand through Amazon or Shopify typically relied on personal intuition, Amazon's bestseller lists, and manual keyword research. The process was slow, expensive when wrong, and heavily weighted toward copying what was already popular rather than finding real gaps in supply.
Tools like Alibaba's Accio - an AI-powered product discovery and sourcing platform - are changing that calculus. Accio lets sellers describe a product category or customer problem and returns supplier matches, market demand signals, and trend data. The pitch is that a single seller working part-time can now do competitive research that previously required a dedicated team or expensive market intelligence software.
The mechanism matters here. These tools typically work by analyzing large datasets of search volume, social media activity, and sales data - looking for categories where demand is growing faster than supply. The goal isn't to surface bestsellers (those are easy to find yourself in five minutes) but to surface what customers want and can't easily find.
The Window Is Short
For small sellers, the competitive edge isn't just finding better products - it's moving faster than larger competitors who have purchasing minimums, approval cycles, and quarterly planning calendars. An individual seller using these tools can identify a trend, contact a manufacturer, and have a product listed within weeks.
The limitations are real. AI sourcing tools are pattern-matching on existing data, which means they're better at identifying trends already forming than genuinely novel opportunities. And if every small seller is running the same tool, the gap fills quickly - early movers capture the value, later entrants compete on price in a crowded market.
There's also a quality problem AI doesn't solve. These tools surface manufacturers, but evaluating factory reliability, minimum order quantities, and lead times still requires direct supplier communication and human judgment.
What's actually shifting is the floor on research quality. A seller who previously skipped market analysis because it was too time-consuming now has a low-friction version available. That raises baseline competition across the board - which means sellers who were already doing rigorous research have to work harder to stay ahead of people who now have access to the same information.