How to Manually Calculate Van Westendorp PSM's Results

This guide describes how to manually calculate the results for Conjointly’s Van Westendorp Price Sensitivity Meter. To calculate the results of your report, you would have to export the raw data from your study. Then, navigate to the Respondents tab in the exported Excel to find the respondents’ answers to the Van Westendorp question.

To have a simpler demonstration of the calculation, we will use the following example of 10 respondents with their corresponding responses:

Calculate and produce results of Van Westendorp PSM

Step 1: Identify and list down all unique price levels from all 4 columns of the raw data (“too cheap”, “cheap”, “expensive”, and “too expensive”).

Step 2: Aggregate the frequency of the price levels mentioned for “too cheap”, “cheap”, “expensive”, and “too expensive” respectively.

Step 3: Convert the frequency to cumulative frequency

• For “too cheap” and “cheap”, we compute the greater than cumulative frequency by adding all the frequencies greater than or equal to the respective price level. It represents the proportion of respondents who perceive a certain price level to be cheap or too cheap.
• For “expensive” and “too expensive”, we compute the less than cumulative frequency by adding all the frequencies less than or equal to the respective price level. It represents the proportion of respondents who perceive a certain price level to be expensive or too expensive.

Step 4: Based on the cumulative frequencies calculated, plot the chart for Van Westendorp PSM and identify the following points of intersection:

• Intersection of “too cheap” and “expensive” is the lower end of range of the reasonable prices, also known as “point of marginal cheapness”.
• Intersection of “cheap” and “too expensive” is the upper end of the range of reasonable prices, also known as “point of marginal expensiveness”.
• Intersection of “too cheap” and “too expensive” is the “optimal price point” .
• Intersection of “cheap” and “expensive” is the “normal price point” or the “indifference price point”.

The acceptable price range is interpreted as the price ranges from the lower end to the upper end of the range above.

Calculate and produce results of Newton, Miller and Smith’s extension

If you enable the Newton, Miller and Smith’s extension (NMS extension) in your Van Westendorp survey, respondents will be asked how likely they would buy the specified product on a 5-point Likert scale at the price they considered “cheap” and “expensive” based on prices that they inputted in the Van Westendorp question.

With the extension enabled, you will have two additional charts for the Van Westendorp output on your online report:

• Approximate price elasticity of demand chart
• Revenue vs. price chart

In order to calculate and produce the outputs above:

Step 1: The responses to the NMS extension range from 0 (Very unlikely to buy) to 4 (Very likely to buy). In order to make use of the rating scores, we will have to convert the scores to purchase likelihood for each respondent:

• Purchase likelihood - Cheap: Divide the NMS extension’s rating score at cheap price by 5, which results in a percentage value range from 0% to 80%
• Purchase likelihood - Expensive: Based on the assumption that the purchase likelihood at expensive price will never exceed the purchase likelihood at cheap price, we have to conditionally compute the scores such that
  • If the rating score at expensive price is higher than cheap price, we take the purchase likelihood of cheap price instead
  • If the rating score at expensive price is lower than or equal to cheap price, we divide the score by 5
• Compute the difference between the purchase likelihood of cheap and expensive price

Step 2: List down all unique price levels, and aggregate the sum of the purchase likelihood of expensive price and difference for each unique price level.

Step 3: Compute the greater than cumulative frequency of purchase likelihoods by adding all the purchase likelihoods greater than or equal to the respective price level for expensive price and difference respectively.

Step 4: Compute the average cumulative purchase likelihood by adding the cumulative purchase likelihood for expensive price and difference together, and then divided by the number of respondents (10 in this example).

Furthermore, Conjointly’s calculation takes into account of the price levels that are drastically higher than the upper end of the acceptable price range determined from Van Westendorp PSM. It will apply a proprietary formula to adjust the average cumulative purchase likelihood for those price levels, so the results will be closer to realistic expectations.

In our example, this step is not needed as the price levels are not drastically higher than the upper end of the acceptable price range ($8.50).

Step 5: Calculate the predicted revenue (assuming 1,000 units offered in the market) by multiplying the price level by the average cumulative purchase likelihood and then 1000, i.e. Predicted revenue = price * average cumulative purchase likelihood * 1000

Step 6: With the information of average cumulative purchase likelihood and predicted revenue available, we can plot the approximate price elasticity of demand chart and approximate revenue vs. price level chart respectively.