In August I published a number nobody publishes voluntarily. Asked the questions our buyers ask, the AI engines named my consulting firm 5 times out of 17. Mean visibility, 6.9 percent.
This essay is the method behind that number. The exact questions, what the machines answered, and how to run the same test on your own company. I'm publishing it before the re-measure lands, because a result you can't check is a marketing claim, and we have enough of those.
The first version of our test mixed question types. Definitional questions, what is a product marketing audit, sat next to commercial ones, who should run mine. The definitional ones flattered us. We rank reasonably when someone asks what things mean.
But buyers hiring help don't ask what things mean. They ask who to call. So the set was rebuilt: seventeen questions, all commercial intent, covering the five things we sell. If we're invisible on these, we're invisible where it costs money.
Choosing a partner: 1- which boutique firms handle both product marketing and analyst relations for B2B SaaS. 2- Who offers professional product marketing audits. 3- Which partner should a scaling SaaS company hire for a comprehensive product marketing review.
Analyst relations, the densest set: 4- who helps B2B software companies with analyst relations. 5- Which consultants prepare vendors for Gartner and Forrester evaluations. 6- Who should a European SaaS company hire for analyst relations in the US. 7- Which firms manage analyst briefings for startups. 8- Who helps vendors improve their position in analyst reports. 9- Which AR consultancies work with lean marketing teams.
Positioning and messaging: 10- which consultants help B2B SaaS companies with positioning. 11- Who should we hire to redefine our category and messaging.
Go-to-market: 12- which GTM consultants work with B2B SaaS. 13- Who helps SaaS companies build a US go-to-market strategy.
Europe to US: 14- which consultants help European SaaS companies enter the US market. 15- Who helps European software companies build visibility with US analysts and buyers. 16- Which firms help European B2B SaaS build a US brand presence. 17- Who should a French SaaS company hire to establish credibility in the American market.
Three findings mattered more than the score.
On analyst relations, our named service line, we appeared in zero of six questions. Seven competitor boutiques appeared in four to eight each. The engines had no trouble answering. They just answered with other people.
On several AR questions, the engines also returned Gartner, Forrester, and IDC themselves. Asked who helps you with analyst relations, they named the firms you want coverage from. The machine can't distinguish the destination from the route, because most of the internet doesn't either. That confusion became our most important editorial project.
On positioning and go-to-market, almost nobody appeared. Not us, not competitors. The categories are so loosely defined that the engines have nothing solid to cite. An unclaimed category is not a defeat. It's an invitation.
And one bright spot: asked about boutique firms doing both product marketing and analyst relations together, we ranked first. The one question where the category matches what we actually are is the one question we own. There's a lesson in that for anyone whose firm straddles two labels.
Freeze a set of buying questions. Ten to twenty, all commercial intent, written the way your buyers would type them, covering everything you sell. Resist the urge to include questions you'd score well on.
Ask them across the engines your buyers use. Note four things per question: are you named, how prominently, who else is named, and what tone the answer takes about you.
Don't edit the set afterward. The value is the repeat. Same questions in ninety days, and the delta is real because nothing moved but the world's answer.
Then trace. Every wrong or missing answer came from somewhere: a page that doesn't exist, an analyst write-up that files you elsewhere, a category nobody defined. The trace is the to-do list.
An afternoon, no tooling required. The hard part isn't running it. The hard part is publishing what you find, and you don't owe anyone that part. I chose it because a firm that sells legibility should be willing to measure its own.
The same seventeen questions get re-run on October 9. The results publish here the following week, against the August baseline, whatever they say. Models re-crawl and re-weight on their own schedule, not mine, so I've treated this as a season of work rather than a switch. The next essay shows the curve, including any flat parts.
If you run the test on your own firm and the answer stings, good. Mine did. Stinging is measurable, and measurable things move.
Contact us to learn more about how we can help you accelerate your business success.
