Fireflies Discovery Calls: The Checklist I Ask It to Score
Sales calls drift. A fixed checklist prompt in Fireflies turns every discovery call transcript into a scorecard, and the gaps show you what your pitch is missing.
Discovery calls fail quietly: pleasant conversation, good vibes, and afterward nobody can name the customer's actual problem, budget, or timeline. My fix is a checklist that Fireflies scores on every call, and the aggregate gap pattern has improved my pitch more than any coaching did.
The checklist I load: pain stated in the customer's own words (quote it), current workaround and its cost, who else is affected, decision process and who decides, timeline with a forcing event, budget range or a stated constraint, and the next step with a date. Eight items, all binary in the transcript: present or absent.
The prompt to Fireflies runs after each call: 'Score the checklist, quote the exact sentence for each item found, and for anything absent write MISSING, not a guess.' The MISSING discipline is the whole system. AI notetakers want to be helpful and will infer budget from vibes; forcing quotes keeps the scorecard evidence, and forcing MISSING keeps the gaps visible.
One score is an opinion; ten scores are a mirror. Mine showed that 'budget' appeared on two of ten calls while I believed I covered it always. The fix was moving the budget question earlier, tied to the cost-of-the-workaround discussion instead of bolted on at the end. Aggregate patterns are the product; single-call scores are just hygiene.
The second-order effect: the quote bank. Every call now deposits the customer's exact phrasings of their problem into one searchable place, and my proposals open with their words instead of mine. Closing rate moved less than the quality of my follow-ups did, which in discovery calls is the same thing wearing different clothes.