Get recommended by AI.
Buyers now build their supplier shortlist in ChatGPT, Perplexity, Copilot and Google’s AI. I measure how AI sees your company, fix what it relies on and track the change every month.
AI doesn’t browse your site. It reads your data.
In B2B, what an assistant says about you depends less on your homepage than on your product data, your specs and what distributors, directories and customers say about you. For manufacturers and distributors, that’s good news: it’s work you can do.
You’re not in the answer.
A buyer asks for a supplier in your category. Three competitors come up. You don’t, and nobody tells you.
Or you’re in it, wrong.
Discontinued products, old pricing, a territory you left years ago. AI repeats what it finds.
Your specs live in PDFs.
AI can’t cite what it can’t read. Neither can the agent comparing suppliers for a buyer.
SEO habits don’t carry over.
Rankings measure pages. AI answers pick names, and often cite sources you don’t control.
Get into the answer.
Four levers, worked in order of impact. Most of the work uses what you already have: your PIM, spec sheets and key pages.
Also: making your catalog and ordering ready for AI-driven buying. B2B Commerce AI →
- Your buyers’ questions, in English and French.The questions that lead to a shortlist, run across ChatGPT, Perplexity, Google AI Overviews, Gemini and Copilot.
- Product data AI can read.Complete attributes, specs out of PDFs, schema.org markup and an llms.txt file.
- Pages that answer.Category and product pages that answer the question directly, with facts AI can quote.
- A presence beyond your site.Distributor listings, directories, reviews and industry sources AI already trusts.
Measure. Fix. Measure again.
No guesswork and no black box. You see where you stand before anything changes, and every month after.
Baseline
I run your buyers’ questions across the major assistants and score where you appear, what they cite and what they get wrong, next to your main competitors.
Fix
A plan ranked by impact and effort, then the work itself, with your team: product data, markup, key pages, profiles. Commerce builds can go to Human After All.
Re-measure
The same questions every month. Share of voice, citations and AI referral traffic, reported against the baseline.
Share of voiceHow often you’re named when buyers ask, compared with competitors.
CitationsWhich of your pages AI relies on, and which outside sources it uses instead.
AccuracyWhether products, pricing, territory and service are described correctly.Some answers shift within weeks. Product data work compounds over months.
Questions about GEO.
Want the short version first? I covered six concrete actions on the B2B & Go podcast (in French).
What is GEO?
Generative engine optimization: making your company and products easy for AI assistants to find, understand and cite, so you appear in the answer, not only in search results.
How is it different from SEO?
SEO earns a ranking on a results page. GEO earns a mention inside an answer. They share foundations, but AI leans harder on structured product data, consistency across sources and third-party mentions. Good SEO helps. It isn’t enough.
How soon will we see a change?
Some answers move within weeks of fixing a key page or profile. Product data and third-party presence take longer and compound. That’s why we measure monthly against the baseline.
Do we need a new platform?
No. Most of the work is in what you already have: your PIM, ERP exports, spec sheets and key pages. If a commerce build is needed, it’s scoped separately.
Does it matter that we sell in French too?
Yes. Assistants answer differently in French and English, and often draw on different sources. We test and fix both.
What does AI say about you today?
Send me your category and three competitors. We’ll start with the questions your buyers actually ask.
