Mention Network
AI visibility solution for e-commerce
A year ago, a shopper found your store through Google. Now a lot of them just ask ChatGPT, Gemini, Claude or Google AI "where can I buy product X", and the answer names a few stores. If yours is not one of them, a rival selling the exact same product is, and you never see it happen.
I kept running into Shopify resellers who were quietly losing sales to this and had no way to check. The AI visibility tools that exist are built for marketers at big brands. They measure whether a brand gets mentioned, they stop at a score, and they are not much use to the person actually running the store.
But "brand mention" and "where to buy" are two different things. A brand can be all over AI answers while its resellers never get named at the moment a shopper is ready to buy. That gap is what I wanted to close.
So I have been building Mention Network, an AI visibility solution for e-commerce, as an embedded Shopify app. It does four things:
- Check: see whether AI names your store when shoppers ask where to buy a product
- Audit: a health score across 40 signals in four areas (discoverability, content, trust, competitiveness)
- Plan: a specific fix written for your store and product
- Deploy: apply the fix in one click inside your Shopify admin
Right now it is launching first as a Shopify app, with WooCommerce and an API on the roadmap. Your first product check and audit are free, no card, and later plans bill through Shopify so there is no separate checkout.
I would love feedback from other founders here: if you run a store, how would you even know today whether AI is recommending you? And if you have built in the "AI visibility" space, what did you get wrong the first time?
6 Comments
Curious if there are any other platforms / options to use it with other than Shopify?
I think you've identified a problem that a lot of store owners probably haven't thought about yet.
Most people can check their Google rankings, but almost nobody knows what ChatGPT or Gemini recommends when someone asks where to buy a product. It's easy to assume everything is fine when there's no obvious way to measure it.
I'm curious—have you found that AI tends to recommend larger retailers by default, or are smaller Shopify stores showing up more often than you'd expected?
This is a smart distinction; most AI visibility tools stop at brand mentions, but the actual purchase moment (which reseller gets named) is what matters for revenue. Curious how the audit weighs 'trust' signals versus 'content' ones
How do you handle cases where a store sells on multiple channels (Shopify + Amazon + a physical location); does Mention Network track visibility across all of them, or just the Shopify storefront?
I like the distinction you're making between being mentioned and being recommended as the place to buy.
Those can look similar on the surface but lead to very different outcomes. Treating them as separate problems makes the product feel much more grounded than another generic AI visibility tool.
Is this something that you check everyday, does it give you a report. Does it add steps to the workflow? I feel most people just want to be doom and gloom about AI and not really try to harness it. So I like that you are thinking of ways to fight back. Although i do feel that we have to be careful of adding more steps in the routine of a normal day. Its a great idea but I'm not seeing how this would fit into my day other than having to remember to go check it. Not a deal breaking step. I think you should lean into integrating this into a normal workday.
Real problem, and it's only going to compound. One nuance worth separating in your health score: is the AI's answer coming from live browsing/search grounding (Gemini's grounding, ChatGPT with browsing on) versus purely from training data? Those behave very differently -- a grounded assistant can pick up a fresh SEO/content fix within days, but a non-grounded one won't reflect it until the next model update, months out. If the 'Deploy' fix promises visibility gains, worth being upfront with customers about which of those two buckets they're actually affecting, so expectations on timeline don't get set by the easy case.