Reflexa for online stores.
A shopper asks an assistant where to buy, and the answer names two or three stores. Reflexa reads your store the way ChatGPT, Claude and Perplexity read it, fixes what keeps them out where it can run inside the site, writes the words you approve — or the exact text to paste where it cannot run — and counts every AI visit to every product page.
The moment that decides it
The shopper never opens a comparison page. The answer is the comparison, and it is built from whatever the engines could read. Questions of this shape, every day:
- “best <product type> under <price>”
- “where to buy <product> that ships to <country>”
- “<brand> vs <brand> for <use>”
- “is <store> legit”
- “most durable <product> according to reviews”
What breaks on a store site
- Is your CDN or bot filter challenging AI crawlers? — Bot management bought to stop scrapers challenges the AI crawlers too. The store is not beaten in the answer, it is absent from it.
- Does your robots.txt let AI crawlers in? — A platform robots.txt that disallows the collection and search paths — exactly the paths a shopping question reads.
- Does your site say who you are in structured data? — Product schema is there, the Organization behind the store is not, so nothing machine-readable says who the seller is.
- Is your text in the HTML, or only after JavaScript runs? — Themes that draw descriptions, variants and prices client-side. The fetcher gets a shell and moves on.
What the agent does about it
- Opens the door: AI crawlers allowed in robots.txt, one switch, reversible.
- Puts Organization data on every page so the store is an entity, not a name — and a sitemap and /llms.txt beside it.
- Writes the answer-first opening and the FAQ for a collection or product page, in your words, published the moment you approve them.
- Counts every AI visit page by page: which product pages an assistant opened to answer a shopper right now, which it only crawled, who clicked through.
Your first week
- What keeps AI out of the store, in the order that matters, with the exact change for each.
- Crawler access, Organization data, sitemap and /llms.txt applied where the agent runs inside the site, written out to paste where it does not.
- The first collection pages rewritten to answer the question a shopper actually typed.
- The first AI visits counted, page by page.
What it costs: the page check is free in a free account; the agent reads your pages from outside and shows what keeps AI out. Putting the agent on the site starts at $49 a month; five times the usage is $149, and sites are not counted. Pricing in full →
Where the agent goes in
Online stores usually run on Shopify, Next.js, Webflow: one guide each, with the exact file and the one-line check that it worked. [All install guides →](/docs/install)
Who else it is for
Questions people ask
Does Reflexa work on Shopify?
The check runs on any store from outside. On Shopify the agent cannot run inside the site, so it writes the exact text and snippets — robots lines, Organization JSON-LD, opening, FAQ — and you paste them in the theme editor; AI visit counting and the one-click switches need the plugin, an edge file or a connected repo. The Shopify guide is the exact steps.
Will it change my product pages without asking?
No. The four structural fixes are switches you press. Words are different: the agent writes them, you read the exact text, and nothing is published until you approve it. Every change is reversible.
We have 4,000 products. Does it check all of them?
The check reads the pages that matter for how an engine understands the store — the homepage, collections, the templates behind product pages — not 4,000 near-identical URLs. AI traffic, on the other hand, is counted on every page an engine actually visits.
Our Trustpilot reviews are good. Is that not enough?
Reviews are a large part of what a shopping answer is built from, and Reflexa checks that your profiles are linked from the site so the engine knows they are yours. But if the crawler cannot read the store itself, the reviews describe a shop the answer cannot point to.
How is this different from SEO?
Most of the work overlaps: a page a crawler can fetch and read serves both. The difference is the destination. Google ranks a list of links; an assistant writes one answer and names a few stores. Reflexa works on the second, and the first improves with it.
Start with the free check. Then tell the agent what you need.
The free check runs the 23 checks on your public pages. Nothing is installed, nothing changes until you press it.
Free check from outside in minutes · the report by email · nothing installed yet