4 Minutes 
Total Unduplicated Reach and Frequency
Picture this scenario: An ice creamery in your neighborhood recently announced they would be celebrating their new freezer by introducing three new flavors and invited their social media followers to vote for the additions. Hooray for everyone, right? Not necessarily. It’s possible that a simple poll among existing customers – fans! – would not lead to more customers, and in fact may cannibalize existing sales if a customer simply swaps out an old favorite with a new one. A poll among residents within the creamery’s trade area is a slightly better option, since it includes potential new customers, but will not ensure that higher awareness will translate into higher volume nor attribute additional sales to the new flavors.
Instead, the ice creamery should defend its turf. T-U-R-F, or Total Unduplicated Reach and Frequency, is a stronger approach to luring new customers with new ice cream flavors while encouraging existing customers to notice the new flavors. Here is how it works.
Ask a sample of the market (not just current customers) what flavors they would buy. The list should include both the existing flavors, and the new flavors being considered (MaxDiff is best, but you can work with multi-select check boxes too). TURF defines thresholds for each person as to whether they would buy each flavor of ice cream. The analysis then runs through all possible subsets of flavors. You want to reach additional people you are not already reaching through your existing flavor selection (i.e., what set of three new flavors will reach the largest number of additional people?). You force all the current flavors into the analysis first to see who is being reached. Then you test all possible subsets of three from the new flavor list and find the optimal combination(s).
Reach is the number of respondents who expressed a preference for at least one of the items in a combination. If at least one of the flavors above a respondent’s purchase threshold is in a particular subset of products, they are considered “reached.” Unduplicated means you are trying to reach as many unique individuals as possible. In other words, once the analysis has reached the respondent by adding a flavor they would buy, it seeks to identify flavors that will reach other people. For example, suppose the first five flavors identified by TURF are Chocolate, Vanilla, Strawberry, Mint Chocolate Chip, and Chocolate Chip Cookie Dough. All the people who reported liking Chocolate Chunk also liked at least one of the five listed above, so adding that flavor does not bring any additional customers in. While it might be a smaller set of people that chose the Non-Dairy Chocolate, these people are not currently being reached by the previous five flavors (unduplicated reach).
Frequency is the total number of times the features in each combination have been preferred. In this example, this would be the number of flavors in each flavor set that people have indicated they like. Assume the final flavor set is Chocolate, Vanilla, Strawberry, Mint Chocolate Chip, Chocolate Chip Cookie Dough, and Non-Dairy Chocolate. If Mary likes both Vanilla and Chocolate Chip Cookie Dough, her frequency would be two. If Juan likes only Non-Dairy Chocolate, his reach frequency would be one. In this scenario, people are likely to only get one ice cream cone during each visit, so frequency is not as relevant as it would be for something like a fast-food restaurant where each customer is likely to pick multiple items per visit, such as a burger AND French fries.
By conducting a quantitative market research survey and executing a TURF analysis, the ice creamery could identify the flavor list that would keep its existing customers happy while attracting the greatest number of new customers.
Interested in learning more about TURF? Reach out at [email protected]
