Claims in this piece are observable and methodological — no client or benchmark data used.
A commercial head loses an enquiry they expected to win. The feedback, if any comes at all, is thin: “we went in another direction.” They pull the lost-deal report, and it tells them little, because the hardest losses are the ones that never reach the report. You cannot analyse a deal you were never in. The question underneath is simple and uncomfortable: when a buyer went looking for a manufacturer like us, were we ever really in the running?
That question, and a handful like it, is what an AI Discovery Audit™ is built to answer. It does not begin with AI jargon. It begins with the questions a commercial or quality leader already asks, and it replaces the guesswork in the answers with observation. Emerivo is a specialist advisory that helps pharmaceutical manufacturers and research organisations understand and improve how AI assistants find, understand and recommend them; the audit is the part that turns those questions into findings.
The questions underneath
The parent piece, What an AI Discovery Audit Actually Examines, sets out the six dimensions the audit measures. This is the other side of the same coin: the plain-language questions those dimensions exist to answer. An audit does not hand a leader a new vocabulary to learn. It takes the questions already being asked in commercial reviews and gives them an answer grounded in what the assistants actually say.
The questions a leader asks — and what the audit checks to answer them
- “When a buyer asks an AI for a manufacturer like us, do we come up at all?” (Presence)
- “Is what the AI says about us even correct?” (Accuracy)
- “Does it describe what we actually specialise in, or a generic version of us?” (Specificity)
- “Does it know enough about us to put us on a serious shortlist?” (Completeness)
- “Why does one source say one thing about us and another say something else?” (Coherence)
- “Can it back up what it claims, or is it guessing on our behalf?” (Supportability)
Each question is answered by the check named beside it. The audit’s work is to run them across the five assistants a buyer might use and report, plainly, where the answer is good and where it is not.
Why are we losing to a competitor we know we’re better than?
This is the question most leaders actually feel, and the answer is often less about capability than it looks. A competitor who is easier for an assistant to read, place and corroborate gets named in the first pass, while you are left out of a shortlist you never knew was forming. You may not be losing on what you can do, but on how legibly you say it. An audit cannot promise you will win the next one; no one can. What it can tell you is whether the reason you lost the last one was something on your own site that you are able to fix.
Of the enquiries we lost last year, how many were we never actually in — and would we even know the difference?
The questions it will not flinch from
A number that flatters you is easy to sell and useless to act on. The questions worth asking are the ones with uncomfortable answers: where are we described wrongly, where are we invisible, where does our own story contradict itself across the places a buyer checks. An audit earns its place by answering those plainly, not by handing you a reassuring score that makes the quarter look better than it was. If a finding is going to cost you a conversation you assumed you would have, you want to know before the buyer does.
Ask the first question yourself
You do not need to wait for an audit to answer the first one. Open ChatGPT and ask it to recommend CDMOs for a capability you genuinely offer, say modified-release oral solids for a regulated export market, without naming your company. Ask Gemini and Perplexity the same, and Claude and Copilot if you have access. If you do not appear, your Presence question is answered. If you appear but the description is wrong, that is Accuracy. Treat it as observation, not proof; answers vary and shift. But it turns an abstract worry into a specific, checkable one in an afternoon.
Where to go from here
These are the questions behind an honest lost-enquiry review, at any TPM, CMO, CDMO or CRO, and they are exactly what an AI Discovery Audit™ is built to answer. For what the audit examines behind each question, read What an AI Discovery Audit Actually Examines.
Frequently asked questions
The common mistake is the vanity prompt, your own company name in the box. Buyers do not search by your name; they search by capability, market and constraint. An audit asks the questions the way a buyer would, which is the only version that tells you anything about how you are actually discovered.
They often are, and that is the point. An audit that only confirms what you hoped is not worth commissioning. The value sits in the gap between how you see your own capability and how an assistant describes it to a buyer who has never met you.
Both. The audit answers the questions above and turns the gaps into an ordered list of what to correct first, cheapest and highest-trust changes ahead of the longer work. What the audit produces in full is set out in the cornerstone.