Value interviews: past spending, not promises

«Would you pay 800 for a lesson?» feels like a quick way to learn willingness to pay. It isn't. It asks people to predict their own future behaviour about a price you control, and people who like you tend to be polite.

What people already did is a fact: what they tried, what they paid, what it cost them when the problem wasn't solved. Asking about that is a value interview.

Compare four questions Dana could ask:

• «What did you try for your English before you came to me, and what did it cost you?» is past and specific: a fact the learner can recall.
• «Would you pay 800?» asks for a prediction.
• «How much would a perfect course be worth to you?» asks for a fantasy.
• «Don't you think my lessons are worth more?» leads to a polite yes.

A value interview asks about three things, all in the past:

1. Alternatives: what else did they use, and what would they do if you stopped tomorrow? That names the next-best alternative.
2. Money and time already spent: courses, apps, other tutors, hours.
3. The cost of the problem: what happened when it went unsolved.

Dana tried it with Marko, a sales manager with weekly calls to partners abroad, without mentioning her price once. In the past year he paid for a group course, 4 months at 2,000 (8,000), five lessons with another tutor at 500 (2,500) and a language app (1,500): 12,000 in all, on things he found didn't work.

He never named a number for Dana's price, yet the notes are full of money. That spending is a fact about his willingness to pay, not a promise.

The interview's alternatives question pays off when you read the answer like a pricer. Marko's history includes a group course, another tutor and an app, but none of those is his next-best alternative: he dropped them all.

He said it himself when asked what he'd do if Dana stopped: «I'd look for someone who does the same, not another course.» So his alternative is another tutor who does role plays built from real work calls, and the market price of that kind of tutor is Dana's reference, the starting point of her ceiling.

Past spending shows the buyer's budget habits; the "if you stopped tomorrow" answer shows the alternative. Keep them apart: money spent on things that didn't work isn't the price of what they'd choose next.

Interview notes mix strong and weak material. Sort every line into one of two piles.

Evidence is what people did or what happened: past spending (Marko paid 2,000 a month for a group course and dropped it after 4 months), the cost of the problem (in August he stayed silent on a client call and his lead took over), and real choices with their reasons (Olena chose a group course at 1,600 a month because it fit her evenings).

Weak material is pleasant: compliments («the best thing I've tried») and predictions («I would probably pay more for a good tutor»).

Two more rules. Don't turn the cost of the problem into a hryvnia figure the learner never gave you: note it as a strong reason and count how many learners tell a similar story. And interview only your own learners and people who came to you, never Klasna's families.

Not every "no" is about price. Before reading a lost learner as a price signal, check the reason they gave.

Olena asked about lessons in September and didn't start. She chose a group course at 1,600 a month because it had a fixed Tuesday slot and Dana's free evenings didn't match hers. Price came up only at the end, and she didn't push back on it. Her reason was the schedule, so her choice says little about whether 600 is right, in either direction.

More generally, one interview is a story; five or six that agree are a pattern. Write each one up the same way, alternatives, money and the cost of the problem, so the pattern becomes visible when you lay them side by side.

An AI tool can turn messy interview notes into a tidy summary in the same three headings, which saves real time. Two conditions:

• Check every line of the summary against what was actually said. A tool can smooth a hesitant answer into a confident one or merge two people's stories.
• Never ask it to play «a typical learner» instead of talking to real ones. A simulated respondent can sound convincing, but it can't tell you what anyone actually paid, tried or lost. That's fiction, not evidence of willingness to pay.

The test is simple: could you point to the person who said it? If not, it doesn't go into the pricing notes.

The three interview questions aren't random: each one feeds a piece of the price corridor.

• «What would you do if I stopped teaching tomorrow?» names the next-best alternative, the starting point of the ceiling.
• «What have you paid for English in the past year?» shows the money already spent, real evidence of willingness to pay.
• «Tell me about the last time your English let you down» shows the cost of the problem, the size of what a solution is worth.

Each asks about a fact the learner can recall from memory, so the answers don't depend on politeness or imagination. Together they give the price conversation a factual base instead of a guess.

Suppose four of Dana's five learners answer «yes» to «Would you pay 800 for a lesson?» It feels like proof. It's a polite prediction from people who like her, possibly asked in front of others, about a price she controls.

Far stronger: what they already paid for alternatives, and what they do when a real price is in front of them. The interviews won't hand Dana a price either. They tell her what she competes with and what the result is worth, so the price conversation starts from facts. If most learners say they came from 500-a-lesson tutors who only drilled grammar, the reference is 500 and her difference is the call practice.

How it fits together: ask about the past, cover alternatives, money spent and the cost of the problem, keep facts apart from compliments, and look for patterns, never simulated answers. The next lesson moves to a survey: Nazar's four price questions for the Club.