When a hospital buyer asks ChatGPT who solves this, your name should come back.
Book a Strategy CallA service line director used to open Google. Increasingly they open a model instead.
They'll ask something like "what are the options for robotic transcranial doppler" and read the summary. Three companies get named. Everyone else is invisible, and there's no page two to be on.
That summary gets assembled from what the model learned about your category. So if it never learned your category exists, you're not in the answer. Same if it never associated your company with it.
Search rewarded whoever published the best page. Answer engines reward whoever the model already associates with the problem, because it is not ranking pages any more. It is writing a sentence and deciding whose name goes in it.
Answer engine optimization is the work of becoming one of those names. That is the whole of it.
And the work that wins here is the work described everywhere else on this site.
A category with a name. Consistent language across every property. Third-party sources repeating it, and proof a model can find and attribute to you.
A company with a defined category and a body of published work gets named. A company with a vague product description doesn't, however good the product is.
There are two reasons, and both are temporary.
Almost no medtech company is doing this. The agencies selling answer engine optimization work in software and ecommerce, where competition is brutal. Your category is close to unclaimed.
And medtech categories are narrow. A model asked about a broad consumer topic has millions of sources to weigh. Asked about robotic transcranial doppler, it has very few. So a modest body of authoritative work moves the answer.
That window closes as your competitors notice.
It isn't a trick. Anyone selling you one is describing something that'll stop working.
There's no prompt to insert, no schema tag that forces a mention, and no way to buy placement. What moves the result is what always has. Being genuinely associated with a category, in public, in consistent language.
We'll tell you if your category is too undefined for this to work yet. That's a medtech branding problem, and it comes first.
Device, diagnostics and digital health companies with a defined category and published work to draw on.
It works best alongside founder-led content and search, because all three feed the same body of published work. Bought on its own it's a measurement exercise with a short list of fixes attached.
We run this on MarketCraft's own properties. Updating how models describe our clients is a standing step in the monthly cycle, not a product we invented for this page.
MarketCraft is a medtech market engineering agency, founded by Omar Khateeb. Omar sold surgical robots into hospitals before he built the firm. He now hosts State of MedTech, a podcast past 400 episodes with the founders, investors and operators in this industry. Several hundred hours of indexed medtech conversation is exactly the kind of material these models learn from.
The method isn't ours alone. MarketCraft is acknowledged in Market Engineering by Bruce Cleveland, published June 2026. Cleveland ran product and marketing at Apple and Oracle, then became a founding partner at Wildcat Venture Partners.
"MarketCraft joined the board of investors meeting, they were very impressed by MarketCraft's results."
Lloyd Mencinger, CEO, Aqua Medical
The inputs overlap and the target differs. Search ranks pages, and answer engines assemble responses from associations. So a page can rank well and still never get cited.
No, and treat anyone who does with suspicion. We can show you where you stand today, move the inputs, and measure the change.
We see movement in a quarter on narrow categories. Broader ones take longer, and we'll tell you which yours is before you commit.
ChatGPT, Claude, Gemini and Perplexity, plus others as buyers start using them.
Then this is early, and defining the category is the higher-value spend. We'd rather sell you that.
We'll run your category questions before the call and show you the answers. Including who gets named instead of you.
Thirty minutes. You keep the audit either way.
P.S. Most medtech founders have never asked a model to describe their own category. It takes two minutes, and it's usually the moment this becomes urgent. Try it before the call and bring what you get.