Skip to main content
“What does this problem cost you today?” is a question almost nobody can answer on the spot. You’ll get a pause, a hedge, and a promise to check with someone. Then you’ll follow up twice and get a number that turns out to be a guess anyway. Try the opposite. Put a number on the table first: “We think this is running you about 1.4Mayear—here′sthearithmetic."Nowthey′renotbeinginterviewed,they′rereviewingyourwork.Andpeoplecorrectwork.They′lltellyoutheheadcountiswrong,thatyou′vedouble−countedtheoffshoreteam,thattherealnumberiscloserto1.4M a year — here's the arithmetic." Now they're not being interviewed, they're reviewing your work. And people correct work. They'll tell you the headcount is wrong, that you've double-counted the offshore team, that the real number is closer to 900k. You’ve just been handed the data you were fishing for, and — this is the part that matters — they now own the number, because they’re the one who fixed it.
1

Build the hypothesis before the call

Get a value story on paper while you still know very little.
2

Put real numbers behind it

Use benchmarks from your own deals, not invented magnitudes.
3

Show the arithmetic, not the answer

A number they can’t inspect is a number they can dismiss.
4

Get corrected

Invite it directly. This is the whole point.
5

Capture the correction where it counts

Their name on the input changes who owns it.
6

Reflect it back

Send them their own words, so the champion can forward it.

​
Step 1: Build the hypothesis before the call

Minoa generates a value hypothesis for deals it has context on — the qualitative story of how you’d create value here, before any numbers get involved. Read it, tighten it, and tailor the language to whoever you’re meeting. Create a Value Hypothesis walks through generating, refining, and exporting one. Do that first; this recipe is about what you do with it in the room.

​
Step 2: Put real numbers behind it

A qualitative story invites agreement. A quantified one invites correction, which is far more useful. Open a use case, click into any input, and look at the Benchmark section. You’ll see:
  • Your current value against a box plot of the distribution — min, 25th percentile, median, 75th, max
  • Filters for industry and company size, with a count on each so you know how thin the slice is
  • The deals behind it — the actual business cases and accounts making up that distribution
  • A suggested value you can apply in one click, showing what it does to the use case total before you commit
These benchmarks come from your own workspace — the business cases your team has built. Not a cross-customer pool, not an industry survey. When a buyer asks “where’s that from?”, the honest answer is “thirty-odd deals we’ve run with companies your size,” and that answer holds up.
An input needs to appear in at least three business cases before a benchmark shows, so newer use cases won’t have one yet. There’s also a toggle to exclude inputs still sitting on their default value — worth turning on, since a default that nobody ever changed isn’t evidence of anything.
Filter to their industry and headcount band before you quote anything. “The median across everyone” is a weaker opening than “the median for manufacturers your size,” and it’s the same two clicks.

​
Step 3: Show the arithmetic, not the answer

A headline number is easy to wave away. A headline number with five visible inputs underneath it is an argument, and arguments can be had. Share the business case with your champion as an Editor before or during the call, so they can change values themselves rather than emailing you corrections. See Sharing and collaboration for roles and link access. Then walk the inputs out loud, in their order of shakiness:
“We’ve assumed 40 people touch this process, at a loaded cost of $95k, spending roughly a fifth of their week on it. That’s the whole model. Which of those three is most wrong?”

​
Step 4: Get corrected

Ask for it plainly. Most buyers won’t volunteer a correction to a vendor’s slide, because correcting it feels like engaging, and engaging feels like buying. Take that hesitation off the table:
“I’d genuinely rather be wrong here than have you nod at a number neither of us believes. What would you change?”
The goal isn’t to be right. It’s to be wrong by a knowable amount. A hypothesis that’s 40% too high and gets fixed in the room is worth more than a cautious estimate nobody engages with.
Two things happen when they correct you. You get a real number. And they stop being an audience.

​
Step 5: Capture the correction where it counts

Where the corrected number lands determines whether it survives contact with their CFO. Change it in the input while you’re on the call, or send a baseline survey so the answers come back written by them. Either way, input ownership puts their avatar on the input — Minoa tracks values that came from an accepted survey response, a direct edit by an external collaborator, or a manual assignment. See Input ownership. That avatar is doing real work. It’s the difference between “the vendor says we spend 900konthis"and"ourVPofOperationssayswespend900k on this" and "our VP of Operations says we spend 900k on this,” and only one of those survives a finance review.
Don’t quietly overwrite their number later because it makes the case weaker. If a correction takes the benefit down, that’s the case getting more defensible, not less. The version that closes is the one they can’t poke a hole in.

​
Step 6: Reflect it back

Within a day, send them the Discovery Summary — what they told you, in their words. It attributes each priority to who said it and roughly when, so it reads as a record rather than a pitch, and your champion can forward it internally without editing anything out. You’ll find it on the deal under Docs → Shareable Templates. Value framing stays light here on purpose; it’s early, and the point is to prove you listened.
You now have a business case with the buyer’s own numbers in it, their name attached to the ones that matter, and a written record they’ve already circulated. That’s a much better starting position than a completed discovery questionnaire.

​
Key takeaways

  • A wrong number gets corrected. A question gets deflected. Lead with the hypothesis.
  • Benchmark from your own deals. Filter to their industry and size before you quote a median.
  • Show every input. A number they can’t inspect is a number they can dismiss.
  • Ownership is the whole game. An input with their avatar on it is evidence; the same input with yours is a claim.
  • Send it back in their words. The Discovery Summary is what your champion forwards.