SkillsSkill 3 of 5
Run the discovery an AI summary can’t
Trace every qualification field to the buyer’s own words, and reach the economic buyer and the trigger before you demo.
Worked example
The AI summary says this deal is qualified. Is it?
A filled field isn’t a confirmed one. Trace each claim to what the buyer actually said before you move the deal.
The question
A Ledgerwick AE had a first call with a 400-person logistics company. Ledgerwick is a fictional accounts-payable automation platform. The AI call summary filled every qualification field, and the AE wants to move the deal to evaluation and book a demo.
Pain: manual invoice matching. Budget: approved. Decision maker: VP Finance. Timeline: this quarter. Recommendation: advance to evaluation.
- Senior AP specialist, the only buyer on the callWe key in about 4,000 invoices a month. Month-end is painful.
- Same speakerI think finance has budget for tools this year. The VP Finance would sign anything like this.
- Same speakerNo date yet. Our controller is looking at options after the audit.
| Deals | Won | |
|---|---|---|
| Met the economic buyer and confirmed a trigger | 18 | 9 |
| Met the economic buyer, no trigger confirmed | 12 | 3 |
| Never met the economic buyer | 30 | 3 |
Invented data. 15 of 60 won, 25%. “Met” means the economic buyer joined a call; a name in a summary doesn’t count. Past deals show association, not cause.
“The AI summary marks budget approved, VP Finance as decision maker and a timeline this quarter. Would you advance the deal to evaluation and book the demo?”
Your turn
Say your answer out loud, or jot notes below. Both work. Or skip to the worked answer.
Not saved by this site. Your browser may restore text when you navigate back.
A strong answer, step by step
Open each step in turn. This is one strong answer, not the only one.
Step 1: Check where each field came from
What would you do here? Open the move you’d make.
Trace each field to the call: budget is ‘I think’, the VP is a guess, and the timeline is ‘no date yet’.
The strong move. It separates what the buyer said from what the AI inferred.
Accept the summary: it filled every field from the recording.
Tempting, but it misses something. Filled isn’t confirmed. One AP specialist’s hedges became ‘approved’ and ‘this quarter’.
Rerun the AI summary with a stricter prompt.
Tempting, but it misses something. A better prompt might hedge more, but no prompt can confirm a budget that nobody on the call owns.
See how a strong candidate says it
“Every field traces to one AP specialist. Budget is ‘I think’, the VP is a guess about who signs, and the timeline is ‘no date yet’. The summary turned hedges into facts.”
Why it works: It separates what the buyer said from what the AI inferred.
Step 2: Use the team’s history to size the gap
What would you do here? Open the move you’d make.
Use the 25% overall win rate as this deal’s odds.
Tempting, but it misses something. The average hides the split: 9 of 18 won where we met the buyer and confirmed a trigger, 3 of 30 where we never met the buyer.
Skip the history; every deal is different.
Tempting, but it misses something. The team’s own 60 deals show how much meeting the buyer has mattered: 50% against 10%.
Our history: 9 of 18 won with the buyer met and a trigger confirmed, 3 of 30 without meeting the buyer. This deal sits with the 30.
The strong move. It uses the team’s own numbers to show why confirming the buyer matters, without claiming a cause.
See how a strong candidate says it
“In our last 60 closed deals we won 9 of 18 where we met the economic buyer and confirmed a trigger, 50%, and 3 of 30 where we never met the buyer, 10%. Today this deal belongs with the 30.”
Why it works: It shows why confirming the buyer matters, in the team’s own numbers, without claiming a cause.
Step 3: Plan the discovery the AI can’t do
What would you do here? Open the move you’d make.
Send a demo invite to the VP Finance directly, citing the summary.
Tempting, but it misses something. Nobody has confirmed the VP owns this. Going around the specialist on a guess risks the one contact you have.
Ask what the audit found, who owns fixing it and what doing nothing costs, then invite the controller to the next call.
The strong move. It turns a pain into a trigger, an owner and a cost, and at 4,000 invoices a month the cost can be sized with them.
Ask the specialist to confirm budget and authority on a checklist by email.
Tempting, but it misses something. A checklist invites the same guesses again. The controller, who is looking at options after the audit, is the person to hear from.
See how a strong candidate says it
“I’d ask what the audit found, who owns fixing it and what happens if nothing changes by year-end, then ask to bring the controller to the next call. At 4,000 invoices a month, we can size the cost of manual matching together.”
Why it works: It turns a pain into a trigger, an owner and a cost the buyer agrees with.
Step 4: Decide what stage it’s really in
What would you do here? Open the move you’d make.
Advance to evaluation; the demo will create urgency.
Tempting, but it misses something. A demo before the trigger is known pitches to a user who can’t buy, and it inflates the forecast.
Keep it in discovery, correct the fields, and book the demo once the controller confirms the problem and a date.
The strong move. It protects the forecast without dropping the deal.
Disqualify it: there’s no budget and no buyer.
Tempting, but it misses something. Too fast. There is a real pain at 4,000 invoices a month and a controller already looking at options.
See how a strong candidate says it
“I wouldn’t advance it yet. I’d keep it in discovery, correct the fields, and book the demo once the controller confirms the problem and a date.”
Why it works: A scoped decision that protects the forecast without dropping the deal.
The answer, in short
- Decision
- Keep the deal in discovery, correct the AI-filled fields, and get the controller on the next call before booking a demo.
- Because
- Every qualification field is a secondhand guess, and past deals where we never met the economic buyer won 10% of the time, against 50% where we met the buyer and confirmed a trigger.
- What would change it
- The controller confirming the audit finding as the trigger, with an owner, a budget and a date, or the specialist turning out to hold the budget for a purchase this size.
Common traps
- Treating a filled field as a confirmed one.
- Taking a user’s guess about who signs as fact.
- Demoing before you know the trigger.
- Reading past win rates as a guarantee.
Check your answer
- Did I trace each field to what the buyer actually said?
- Did I use the team’s own win rates?
- Did I plan questions that reach the buyer and the trigger?
- Did I decide the stage and the next step?
Want feedback on how you’d answer? Practice “Run the discovery an AI summary can’t” with an AI interviewer on Socratify.
Last reviewed Nov 9, 2026