Guide

Prompted and unprompted feedback: counting after The Mom Test.

Unprompted feedback is a problem the customer raised before you mentioned it. Prompted feedback is an answer to your question about it. Keep the two apart when you count: a problem many customers raise on their own is far stronger evidence than one they agreed with when asked.

Updated

  • Raised it themselves“Mondays I go through every report by hand.”
  • Agreed when asked“Yes, that would be useful.”

What do prompted and unprompted mean?

The same problem can reach you in two ways. Here are two made-up moments from two calls:

Raised it themselvesAgreed when asked
You: "How has the last month been?"
Them: "Fine, apart from Mondays. I go through every report by hand before the leadership meeting."
You: "Would it help if reports showed what changed since last week?"
Them: "Yes, that would be useful."

Both are about the same problem. Only the first tells you it is on their mind. The second tells you they are polite.

Why does it matter for counting?

The Mom Test by Rob Fitzpatrick is the book most product people read on how to ask customers questions that give honest answers. Its concern is the conversation. This guide is about what happens afterwards, when you add up what you heard.

The trouble is that a list of feedback looks the same whether people raised something or merely agreed to it. "Nine customers want reports to show what changed" might mean nine people brought it up, or that someone on your team asked nine people a leading question. Counted together, a pitch and a problem look identical.

Keeping the split costs one extra column. Leaving it out can make your most-asked-about idea look like your customers' biggest problem.

How do you tell them apart in a transcript?

SituationCount it as
They raise it before anyone mentions the topicRaised it themselves
You ask an open question ("What is hardest about Mondays?") and they pick this problemRaised it themselves. They chose it.
You ask about this topic or a fix by name, and they agreeAgreed when asked
They raised it, then you asked follow-up questions about itRaised it themselves. The follow-ups add detail, not a second mention.
They react to a demo or a pitchAgreed when asked, however keen they sound
They bring it up again on a later call without being askedStill one company. Raised it themselves if they ever did.

The question to ask of every sentence: had anyone from your side said this topic out loud before they did? If not, they raised it themselves.

Which questions let customers raise things themselves?

Questions about their recent past and their own words. Questions that name your idea turn every answer into agreement.

Leads themLeaves it to them
"Would a changes view on reports help?""Walk me through last Monday morning."
"Is reporting a pain for you?""What takes longer than it should?"
"Would you pay for that?""What do you do about it today, and what does that cost you?"
"Do you like the new dashboard?""When did you last open the dashboard, and why?"

You will still ask about your idea at some point. That is fine. Just mark what came after as asked.

How do you count them?

Per problem, keep two numbers: how many companies raised it, and how many of them raised it without being asked. Count a company as raising it themselves if anyone from it did so at least once.

Count companies, not mentions, for both numbers. One customer who raises a problem on four calls is one company with one problem. The guide How to analyse customer feedback from your calls shows the full sheet.

What does it change?

A made-up quarter, two problems:

ProblemCompaniesWithout being asked
Reports do not show what changed92
Nobody hears back after reporting something66

The first looks bigger. But seven of its nine came from a quarter where the team was testing the changes-view idea on calls. The second was raised by every company on its own. If you can only do one, the second is the safer bet.

We stopped reporting things. It goes into a hole.
Caldera Retail, renewal call, raised without being asked (example)

What are the common mistakes?

  • Throwing away what they agreed to when asked. It still tells you how widely a problem is recognised once someone names it. It just cannot tell you how much it hurts.
  • Counting enthusiasm in a demo as demand. People are kind to the person showing them something.
  • Letting a research push skew the quarter. When the team asks about one idea on every call, the number of companies agreeing to it goes up by design. Compare how many raised it themselves instead.
  • Marking it later from memory. Mark it from the transcript. Memory reliably forgets who said the topic first.

Can the split be kept for you?

Marking every sentence by hand means reading what came before it, every time. Most teams stop doing it after a few weeks.

Ordagrann reads the customer calls you record in Granola, groups every problem and request into themes, counts how many companies raised each one, and turns a theme into a one-page brief your team can build from. Every theme shows how many companies raised it and how many did so without being asked, and each count opens to the exact sentences.

Get started

Works with Granola Business. Any other transcript can be pasted in.