Price Testing – aka which price is right

How as market researchers can we help our brethren in marketing, sales and product development assess whether or not our price in the ballpark? There are numerous ways to test price including market experiments designed to measure elasticity up through several survey-based techniques. A key question to review up front is the nature of the product or service. If the product or service is in the concept stage then a choice modeling approach such as conjoint analysis is appropriate.

Choice modeling allows the researcher to test various combinations of feature and benefits of which price is typically a major factor. This approach requires extensive experimental design and programming expertise, and is often sent to a specialist for design and deployment. The key benefit of the method is that it provides a picture of the market’s willingness to make tradeoffs between various features and benefit combinations at different price points. This effectively informs the product design team as to what constitutes an “ideal” product or service.

If the product or service is already in existence then other methods make more sense to employ. The two primary methods include monadic price testing and price laddering. In both cases, respondents are asked to assess their likelihood to purchase at a given price point. This likelihood measure is typically a seven to ten-point measure. Analysis focuses on the respondents in the top-two box (percent most likely to purchase).

In monadic testing the sample is divided into cells each receiving a specified price. This method requires a large sample to draw from in order to prevent margin of error concerns. The more price points the more cells and sample needed. This has cost implications if outside sample is being utilized.

In a price laddering scenario respondents are shown a series of prices, from high to low. If they are not in the top-two box (most likely to purchase), they are shown a lower price and asked again if they would be willing to purchase at the lower price. This continues for all price points. If 100 respondents are shown the first price and 15 are willing to purchase then the “take rate” for the first price is 15%. If 20 of the remaining 85 are willing to take at the second price is 35% (35/100).

Both monadic testing and price laddering will get you close to the truth. Monadic testing tends to provide the purest response as they only see a single price and are not given the option to change their mind. However, this needs to be weighed against the higher cost for split samples. Price laddering tends to lead to a slightly higher take rate, but based on experimental work is not far off from the results displayed by monadic. This can be adjusted downward if necessary.

Photo Credit: “Cash in black wallet” by Stuart Miles at http://www.freedigitalphotos.net/