How to use the gabor-granger demand curve calculator
- Run your survey so that each person is asked, at one or more prices, whether they would buy. Count at each price how many were asked and how many said yes.
- Enter one line per price: the price, the number asked and the number who said yes, separated by commas. Prices can be in any order.
- Enter your cost per unit if you want profit as well as revenue, and choose whether to smooth the answers so demand never rises with price.
- Read the price with the highest revenue and the price with the highest profit, and the table of demand at each price.
Formula
share who would buy at price p = yes count / number asked
smoothed shares: pool neighboring prices until the shares never rise with price (weighted by the number asked)
revenue for each person asked = price x share
profit for each person asked = (price - cost) x share
The smoothing is the standard pool adjacent violators method. Where a higher price shows a higher share by chance, the two prices are replaced by the weighted average of their shares.
Six tested prices
Of 120 people asked at each price, 96, 85, 88, 52, 31 and 30 said yes at $10, $12, $14, $16, $18 and $20. The raw shares are 80.0, 70.8, 73.3, 43.3, 25.8 and 25.0 percent. The share at $14 is higher than at $12, which cannot be true for the same group, so the two prices are pooled to (85 + 88) / 240 = 72.1 percent each.
With a cost of $5, revenue for each person asked is $8.00, $8.65, $10.09, $6.93, $4.65 and $5.00, and profit is $4.00, $5.05, $6.49, $4.77, $3.36 and $3.75. The best price for both revenue and profit among the tested ones is $14.
How a Gabor-Granger survey works
The method asks people whether they would buy a product at a given price and, if they answer yes, may ask again at a higher price. The shares who say yes at each price trace a demand curve. Multiplying price by that share gives the revenue for each person asked, and subtracting cost first gives profit.
The method says nothing about prices you did not test. The best tested price may not be the best possible price, so include a range that brackets what you expect to charge. The Price Elasticity of Demand Calculator measures how sensitive demand is between two prices.
Limits of stated intent
What people say they will pay in a survey is usually more than what they pay in a shop, and a small sample gives a rough curve. A difference of a few answers can move the best price. Use the Sample Size Calculator to plan how many people to ask, and treat the result as one input beside cost, competition and positioning, such as the margin you get from the Cost-Plus Pricing Calculator.
Assumptions and limits
- Each line holds the number of people asked at one price and how many of them said yes. Counts are whole numbers.
- Smoothing makes the shares non-increasing in price. Turn it off to use the raw shares.
- Revenue and profit are per person asked, not per customer, and use the shares you entered or the smoothed ones.
- From 2 to 30 prices are accepted, and each price can appear once.
- Nothing typed here leaves your browser.
Frequently asked questions
Why does the page change my percentages?
If a higher price shows a higher share of buyers, the data contradict the idea that people buy less as the price rises. With smoothing on, those prices are pooled to remove the contradiction. Turn it off to see the raw numbers.
Can the best price be one I did not test?
It might be, but the data cannot say. The page only compares the prices you tested. To look between them, run another round of the survey with prices in the gap.
What if I asked a different number of people at each price?
That works. The shares are computed per price, and the smoothing weights each price by the number of people asked at it.
Why use revenue and profit both?
Revenue ignores what each unit costs you, so it can favor a price that sells many units at a thin margin. Profit uses your cost and usually favors a somewhat higher price.