At TRC, the most popular spot in the office is our snack shelf. It features an array of sugary, salty and carb heavy treats. The contents vary and are determined by one person (Ruth, who stocks the shelf) with influence from the rest of us (based on past usage and suggestions). Sometimes the shelf has exactly what you’re looking for. Other times, not so much. But what if instead of relying on Ruth’s powers of deduction we were to use research to figure out the optimal shelf configuration? We’re researchers, after all.
We would start out by using our Idea Mill™ product to generate ideas on which snacks people want to have. It uses incentive alignment and gamification to bring out the most creative ideas and provide direction on the favorites. It is likely that this will create too long a list of ideas (the candy shelf is only so large) and while we can toss out ideas that are not feasible, we believe it is best not to toss out ideas just because you personally don’t like them (I’m looking at you Mr. Goodbar). Far better to get more consumer input…this time to narrow the list.
We could ask our folks to rate all the suggested snacks and then use that to figure out which ones should make the cut. Ratings might be good enough to eliminate some things (my guess is that despite what people claim, healthy snacks would bite the dust), but among popular snacks (like different types of pretzels) we are not likely to see clear differentiation.
A choice method like Max-Diff could help but if the list was long it would require a lot of work on the part of our employee respondents. A method like our proprietary Bracket™ would do the job in a faster and more engaging fashion while still finding clear winners and losers.
Stocking the winners would therefore make the most sense…but would it please the most people?
Currently the shelf features five types of M&M’s (original, almond, caramel, dark and strawberry nut). If dark chocolate was the least preferred it might get cut. But what if those who like almond, caramel and strawberry nut also liked original, but those who like dark only liked it. For situations like this we can take the results of the Bracket™ (or Max-Diff) and use TURF to find the combination that would please the most people.
Of course, another factor is positioning. The shelf is only so large. M&M’s can be dispensed from any size canister (in fact Ruth has one that spins so that it can dispense three types) while Pretzels tend to come in large bins that take up a lot of room. In addition, not all of the snacks cost the same. In an effort to keep our expenses and waistline under control we follow a strict budget. Might I trade off having greater quantity of a lesser snack in exchange for an expensive favorite?
For these kinds of questions a discrete choice conjoint is the answer. We can include a variety of candy types and constraints related to the room they take up as well as cost. Simulations can then optimize how to spend our candy budget.
Despite our love of research and wide array of tools though, I think in this case they would be overkill (we have a very small population of around 40 employees). So I think we’ll stick with Ruth’s instincts. I never go wanting….