Willingness to Pay
Date: 2026-08-16
The most someone would pay before walking away. It’s the theoretical basis for pricing and it can’t be measured directly — every method of asking produces an overstatement, and the only reliable evidence is behaviour.
What it is
Willingness to pay (WTP) is the maximum price at which a given customer would still buy. It varies by person, by moment and by context.
The gap between WTP and the price charged is consumer surplus — value the customer captures rather than you. Pricing strategy is largely about how much of that gap to close, and how much to leave as a reason to come back.
Stated WTP is unreliable
Asking people what they’d pay produces numbers that don’t survive contact with a checkout:
- Hypothetical bias. No money changes hands, so answers are generous
- Social desirability. People don’t want to seem cheap, or seem extravagant
- No opportunity cost. In a survey, the £40 isn’t competing with anything else they’d buy
- They don’t know. People discover what they’d pay by being asked to pay it
Typical overstatement is substantial, and it isn’t a consistent multiple you can correct for. Treat any survey-derived WTP figure as an upper bound at best — Surveys.
Methods, and what each is worth
| Method | Gives | Reliability |
|---|---|---|
| Direct question | A number | Poor |
| Van Westendorp | An acceptable price range | Better — comparative rather than absolute |
| Conjoint analysis | Trade-offs between features and price | Good for relative value, still stated |
| Observed price changes | Actual response | Good, confounded |
| A/B price test | Causal | Best, and hardest — Price Testing |
Van Westendorp asks four questions — at what price is it too cheap to trust, a bargain, getting expensive, too expensive — and reads the intersections. Better than a direct question because it’s comparative, still stated rather than revealed.
Conjoint forces trade-offs between attributes including price, which is closer to a real decision. Good at telling you what a feature is worth relative to price; less good at absolute levels.
The revealed alternative
The most reliable signals cost nothing to collect and are usually sitting in your own data:
- What people already pay. Your price distribution across the range is revealed WTP for your actual customers
- Response to past price changes — Price Elasticity
- Discount sensitivity. How much volume moves at each depth tells you where WTP clusters — Discount Impact on Margin
- Abandonment at the point price becomes clear — including delivery cost at checkout, which is a WTP signal about the total, not the item
- What sells at the top of the range. If the premium line never sells, you’ve found a ceiling
Segments, not a number
WTP varies enormously across customers, and the practical response is structural rather than analytical: offer a range that lets people self-select.
good £30 for the price-sensitive
better £55 the volume line
best £95 captures high WTP, and anchors the middle
That captures more of the distribution than any single price can, without differential pricing and its problems — Tiered Pricing, Price Anchoring.
The practical position
For most retail decisions, WTP is not measured — it’s inferred from margin requirements, competitive position and break-even arithmetic.
Which is fine. The break-even calculation for a price change requires no WTP estimate at all: compute the volume loss you could absorb, and judge whether it’s plausible. That’s usually a more defensible route than a stated-preference study — Break-Even Analysis, Price Elasticity.