B2B · Software
A software company finds out what AI says about it
A modelled scenario of a B2B company measuring and improving how AI answers describe it.
Scenario
Context
A software company selling to mid-sized firms. Sales said prospects increasingly arrived saying 'I asked ChatGPT'. Marketing had no idea what those tools were saying.
Cause
The problem
Nobody knew whether the company was in the answers.
The product facts differed between the website, directories and review sites. Several pages that explained the product were thin, and one important page blocked AI crawlers by mistake.
What we did
Week 1
Questions
We agree thirty buyer questions that matter, in the buyer's own words.
Week 2
Test
We ask each question across the main AI answer engines and log who is named and what is cited.
Months 2 to 4
Improve
Product facts made consistent, thin pages rewritten, crawler access fixed.
The modelled outcome
In the model, progress is bumpy. That is why we report a pattern over time, not one result.
Illustrative quote
"I didn't expect a straight answer about AI search. The page that listed what they couldn't promise is what made me trust the rest."