App Pricing Research: Choosing a Method and Reading the Results
Most subscription app prices are set by instinct, by copying a competitor, or by working backward from costs. Each approach answers a different question than the one that matters: what will a target user actually pay for the value they perceive? This guide compares three survey-based methods, shows how to prepare respondents, and explains how to read the output before you commit to a price.
When pricing research fits a subscription app
Survey-based pricing work suits most subscription apps, whether already live or still in development. The clear exception is a product whose value depends heavily on brand and long-form storytelling. A premium learning platform, for example, may need extended exposure to convey its range of experts and courses; a short survey cannot reproduce that context. If your value proposition is hard to compress into a 30-second pitch, treat survey results as weaker evidence.
There is also an internal risk to plan for. Research can return a price lower than the team hoped, or below the level needed for profitability. Agree in advance that a low result is information about perceived value, not an automatic instruction to cut price. The first response may be to improve how value is communicated or to strengthen the brand, then revisit the number.
Who to ask and what to ask alongside price
All three methods are surveys, so the sample matters as much as the questions. Recruit people in your target demographic. For an existing app, non-customers are often preferable because they are not anchored by your current price; existing customers make sense when you are testing a new app or a different subscription type.
Sample size guidance: at least 100 responses for directional reading, and 300 or more when the decision is major.
Price answers alone are easy to misread. Add context questions so you can segment the results:
- Operating system. Pooling iOS and Android can produce a misleading conclusion that you are simultaneously too expensive on one platform and too cheap on the other.
- Jobs To Be Done. Ask what the respondent would use the app for. For non-customers, a screening question upfront can filter for relevance. A financial planning app, for instance, might ask how the person currently manages personal finances, with options ranging from active budgeting in an app or spreadsheet to no structured management; those who already budget are the most relevant group.
- Prior spending. Ask whether they have paid for a similar app, and which apps they have paid for, to gauge how realistic their expectations are.
- Current alternatives. Ask what solution they use today. This reveals the real competition, which is often not another app in your category.
- Location. Price expectations vary by country and region. Use a dropdown to keep answers clean.
With a large enough sample, differences in preference across these segments are often more useful than the headline price.
Van Westendorp Pricing Model: four questions, one range
The Van Westendorp model asks four price questions about a single product concept, which forces respondents to consider the price from more than one angle. Begin with a short description of the app: what it helps users do and the benefits, plus a visual if branding affects perception.
The four questions:
- Too cheap: at what price would the product seem so inexpensive that you would question its quality?
- Not a bargain: at what price would it start to feel like a bargain?
- Not expensive: at what price would it begin to seem expensive?
- Too expensive: at what price would it be too expensive?
Responses are recorded per participant with context fields. A sample results layout appears in the pricing research response table below; it shows how operating system and use case sit alongside the four price answers.
[table0001]
To build the chart, plot the cumulative percentage of respondents at each price for each question. The lowest price sits at 100% on the too-cheap curve because everyone would accept it. Four curves result:
- Too Cheap: share who consider the price too cheap or lower.
- Not a Bargain: share who consider the price a bargain or lower.
- Not Expensive: share who consider the price expensive or higher.
- Too Expensive: share who consider the price too expensive or higher.
The intersections give four readings:
- Optimal Price Point (OPP): where Too Expensive and Too Cheap cross. The price that maximizes appeal without damaging perceived value.
- Indifference Price Point (IPP): where Not Expensive and Not a Bargain cross. Roughly equal shares see the price as fair versus too high.
- Point of Marginal Cheapness (PMC): where Too Cheap and Not a Bargain cross. The lowest price at which value is still perceived.
- Range of Acceptable: between Too Cheap and Too Expensive. A price inside this band is unlikely to put customers off.
The OPP is a recommendation, not a verdict. The full diamond can be used to balance profitability against demand, and the acceptable range is a natural input for the Gabor-Granger method described below. If the OPP lands lower than needed, revisit communication and brand before cutting price.
Two setup traps in the Van Westendorp survey
The diamond only forms if the four answers are logically ordered. If a respondent can enter a lower figure for too expensive than for too cheap, the chart breaks. Prevent it with validation rules in the survey tool, or filter those responses afterward on the assumption that the respondent was not paying attention.
Second, a slider or fixed scale makes answering easier but can anchor people to the range you display. Choose the bounds carefully, and if anchoring is a concern, use an open text field instead.
Strengths and limits of Van Westendorp
The method is quick to set up and returns a range rather than a single number, which makes it a practical first look at price perception. Its limits are shared with most survey approaches: little context, and answers based on stated perception rather than real purchase behavior. It also covers one option at a time, so testing several packages makes the survey long. For multi-tier packaging, a different method is usually a better fit.
Conjoint Analysis: comparing packages
Conjoint Analysis splits a product into attributes such as features, price and billing frequency, then asks respondents to choose between packages. The output shows which attributes drive preference and which package different segments would pick. It suits apps sold in tiers more than a single monthly or annual price, because it approximates a purchase decision.
Setup: define three or four packages, each with a price and its key features, and present them as choices. Keep the number of packages low; too many options can produce weak selections, so four is a sensible ceiling. Because the method assumes you already have an approximate price in mind, it is a poor choice for first-pass pricing. Like the other methods, it may miss brand perception, external influences and situational factors.
Gabor-Granger Pricing Method: measuring price sensitivity
Gabor-Granger focuses on price elasticity: how much demand changes as price changes. High elasticity means discounts and offers tend to work well, while raising prices is harder. Low elasticity suggests the reverse.
The steps:
- Pick a range of price points.
- Present the product in a survey.
- Ask whether the respondent would buy at each price.
- Find the price with the highest share of willing buyers.
On the resulting chart, apps with high elasticity show a steeper drop-off as price rises. A variant called Price Laddering works well as a follow-up to Van Westendorp: start at the OPP, ask whether the respondent would pay it, then move down if the answer is no or up if it is yes, repeating until you reach the highest price they accept.
The method is fast and needs little resource, which makes it useful for new apps without sales history or clear competitors. Its weakness is the same stated-intention gap as the other methods, plus an assumption of a linear relationship between price and demand. It does not account for competitor pricing, changing preferences or market shifts.
Choosing between the three methods
Use the comparison below to match the method to the question you need answered. The same table is available as a structured reference in the pricing method comparison table.
[table0002]
A practical sequence for many apps: run Van Westendorp first to establish a perceived-value range, follow with Gabor-Granger or Price Laddering to test sensitivity inside that range, and use Conjoint when the decision is about how to package features across tiers. Combining two methods is often more informative than relying on one.
From research to a price you can test
Research narrows the starting point; it does not finish the job. Treat the output as a hypothesis and validate it with a controlled experiment. Keep experiments deliberate: frequent price tests can create inconsistent experiences and negative feedback, which makes the results hard to trust.
Before launching, confirm four things:
- The sample matches your target demographic and is large enough for the decision.
- Platform and use-case segments are separated in the analysis.
- The chosen price sits inside the acceptable range, or you have a documented reason to go outside it.
- A plan exists for what happens if the test underperforms, including whether the response is a price change or a value-communication change.
A price set this way is a considered starting point, not a permanent answer.
| Participant | Operating System | Relevant Context | Too Cheap | Not a Bargain | Not Expensive | Too Expensive | |
|---|---|---|---|---|---|---|---|
| 1 | John Doe | iOS | Use case 1 | $40 | $60 | $100 | $120 |
| 2 | Jane Doe | Android | Use case 2 | $20 | $35 | $45 | $55 |
| 3 |
Illustrative sample rows from the original article; not market data.
| Pricing Method | How It Works | Best Use Case |
|---|---|---|
| Van Westendorp Pricing Model | Asks four key questions to determine a price range based on perceived value: too cheap, bargain, expensive, and too expensive. Creates a price range based on customer perception. | Best for understanding price perception and acceptable price ranges for an app. Ideal for early-stage pricing or repositioning an existing app. |
| Conjoint Analysis | Breaks down a product into attributes (e.g., price, features, billing frequency) and asks respondents to choose between different packages to understand how they prioritize attributes. | Best for apps with multiple pricing tiers or feature-based plans. It helps determine which combinations of features and pricing maximize appeal. |
| Gabor-Granger Pricing Method | Tests a range of price points by asking users if they would purchase at each level to determine demand elasticity and optimal price. | Best for understanding price sensitivity and price elasticity. Useful for gauging the impact of price increases or discounts. |
Method descriptions as presented in the original article.
Frequently asked questions
Have you paid for a similar app before?
This question gauges pricing expectations and how realistic they are. It also shows whether respondents have actually spent money on a comparable product.
Which of these apps have you used and paid for before?
It helps reveal what is shaping the respondent's perspective and how much they have spent on competitors in the past.
What solution(s) are you currently using to solve this problem?
This question clarifies the respondent's Jobs To Be Done and identifies the real competition, which is often not what you expect. It can serve as an alternative to asking which apps they have paid for.
Where are you based?
Pricing expectations differ significantly by country and even region. You cannot set a different price for every location, but this context helps interpret the results. Use a dropdown to avoid messy answers.
How many survey responses do I need for pricing research?
At least 100 responses provide directional guidance. For a major pricing decision, aim for 300 or more.
What if the Van Westendorp Optimal Price Point is lower than I need?
A low OPP may indicate that perceived value is lower than expected, or that there is a gap in how value is communicated. Before dropping the price, consider improving communication or strengthening the brand.
| Participant | Operating System | Relevant Context | Too Cheap | Not a Bargain | Not Expensive | Too Expensive | |
|---|---|---|---|---|---|---|---|
| 1 | John Doe | iOS | Use case 1 | $40 | $60 | $100 | $120 |
| 2 | Jane Doe | Android | Use case 2 | $20 | $35 | $45 | $55 |
| 3 |