What is Gabor-Granger?
Gabor-Granger is a classic pricing research method that helps you find the price your customers are actually willing to pay.
Instead of asking respondents to name a price themselves, a Gabor-Granger question shows them a specific price and asks how likely they are to buy at that price. Based on their answer, the next price shown is adjusted up or down. This continues across several price points, building a picture of purchase intent across your whole price range.
The result is two views on the same data: a demand curve, showing what percentage of people would buy at each price, and a revenue curve, showing which price maximises revenue. Together, they point to your optimal price point, the price where you get the best balance of sales volume and revenue.
Setting up a Gabor-Granger question
Gabor-Granger sits under the Methodologies category.
To set one up:
Open the question type picker and select Gabor-Granger from Methodologies.
Enter your question title, or use the default: "Would you buy [product] at this price?"
Select the currency for your price points.
Enter your price points manually, one at a time. You can add between 4 and 12 prices, in any order, they'll be sorted automatically from lowest to highest. Each price must be unique.
Optionally, add a message card with a product name, description, and image, so respondents have the context they need before answering.
š” We recommend starting with a qualifying question to make sure you're only asking pricing questions to respondents who'd realistically consider buying. We also recommend adding a message card with an image, so respondents have clear visual context for what they're pricing before they see the price points.
Testing multiple markets? You can add a separate, localised Gabor-Granger question for each market or language. Each one can have its own price list and currency.
Choosing your price points
This is the single most important decision in a Gabor-Granger study, and the most common place these studies go wrong. Your results can only ever be as good as the range you test.
ā
1. Anchor around a realistic price. Start with the price you're actually considering ā your current price, a competitor's price, or your intended launch price.
ā
2. Extend roughly 30ā50% either side of it. If you're considering Ā£5.00, a range of roughly Ā£3.50 to Ā£7.00 gives you room to see the curve turn.
ā
3. Space your points evenly. Clustering prices tightly together (Ā£4.90, Ā£4.95, Ā£5.00) wastes price points and produces a flat, uninformative curve.
ā
4. Go higher than feels comfortable. If everybody says they'd buy at your top price, you've learned that your ceiling is somewhere above your range, but not where. It's far better to include one or two prices that feel too expensive.
ā
āMatch how the product is actually sold. If your category prices at Ā£4.99 rather than Ā£5.00, test the prices people would really see on shelf.
Sample size guidance
As a starting point, we'd suggest a minimum of 300ā400 respondents for a reliable demand curve. If you plan to compare segments ā by market, buyer frequency or demographic ā aim for at least 100ā150 respondents in each segment you want to read separately.
Use case | Suggested sample size |
Directional read (early-stage concept, internal sense-check) | 250ā500 |
Confident pricing decision (launch price, planned price change) | 501ā2,000 |
Major strategic decision (portfolio pricing, multi-market rollout) | 2,001ā4,000 |
What respondents see
Respondents answer on a 5-point scale, from Definitely would buy to Definitely would not buy, at each price point they're shown. The price is shown in bold within the question text.
You can edit the wording of these five answer options to suit your product or tone. Whatever wording you use, the top two answers always move the respondent to the next higher price, and the bottom three always move them to the next lower price, so the sequencing logic stays consistent.
The sequence works like this:
The respondent starts at a randomly selected price point.
If they say they'd buy, they're shown a higher price next.
If they say they wouldn't buy, they're shown a lower price next.
This repeats until their approximate ceiling price is identified.
Respondents can't skip a Gabor-Granger question. If they leave partway through and return later, they'll need to restart the sequence from the beginning. This keeps the dataset complete and comparable across respondents.
Reading your results
Your results view shows:
Demand curve: the cumulative percentage of respondents willing to buy at each price point, as price increases.
Revenue curve: price multiplied by percentage willing to buy at each point.
Willingness-to-buy distribution: a chart showing the % of respondents whose highest price they said they would buy is at each price point.
The key findings board automatically surfaces your optimal price point, along with revenue and demand insights and segment breakdowns (for example, frequent vs. infrequent buyers).
As with other question types, you can:
Apply filters and splits by segment, audience, or wave
View statistical significance on split results
Export your data to Excel
Ask Compass a question about the chart, for example, "Where does revenue peak?"
Add the chart to a board
When a filter is applied, you'll view either the demand curve or the revenue curve at a time, not both together.
š” If you have access to the MCP, you can also visualise the Gabor Granger results across surveys, brands and audience splits.
What the numbers can and can't tell you
Two things are worth understanding before you take a number into a pricing meeting.
The revenue curve is an index, not a forecast. It's price multiplied by claimed willingness to buy, so it tells you which price relatively generates the most revenue. It doesn't account for your costs or margin, which means the optimal price shown is the revenue-maximising price, not necessarily the profit-maximising one. If your cost of goods is high, the price you should charge may well be above the peak of the curve.
Claimed intent to buy tends to be higher than the real purchase intent (people are more generous with hypothetical money than real money). Treat absolute percentages as a ceiling, not a forecast, and focus on where the curve bends rather than the numbers themselves
FAQs
How many price points can I test?
How many price points can I test?
Between 4 and 12. We recommend spacing them meaningfully across the range you want to test, rather than clustering them close together.
Can I test different prices in different markets?
Can I test different prices in different markets?
Yes. Add a separate Gabor-Granger question for each market or language, each with its own currency and price list, and then set the display conditions.
What happens if a respondent drops out partway through?
What happens if a respondent drops out partway through?
They'll need to restart the price sequence from the beginning if they come back. Gabor-Granger questions can't be skipped, to keep the data complete for the demand and revenue curves.
How is the optimal price calculated?
How is the optimal price calculated?
We multiply each price by the percentage of respondents willing to buy at that price, giving a revenue index. The price with the highest revenue index is your optimal price point.
Can I export my Gabor-Granger results?
Can I export my Gabor-Granger results?
Yes, to Excel and PowerPoint
Can I change what price point users see first?
Can I change what price point users see first?
Currently users will see a random price point at first.
Are Gabor-granger questions charged differently?
Are Gabor-granger questions charged differently?
A flat fee of 10 credits per Gabor Granger Question