Detailed consumer testing essential for products to pass the dunking test

The announcement that United Biscuits is to revert to a former recipe for its digestive biscuits because some consumers had missed, what the company called, the ‘taste, texture and dunkability’ of the original recipe, highlights the importance of undertaking detailed and repeated consumer research before fundamental changes to favourite products are made.

Preparation for any change in recipe, though, should include detailed and repeated consumer testing to ensure that the new product meets the same consumer expectations.

A lot of the time research will need to consider the motivation for the reformulation.It is important to generate the right expectations so that you get an accurate reflection of how consumers will react in the real world. This is why some testing is undertaken blind and some branded. Sometimes, though, the questions will also be dictated by whether the reformulation is being driven by internal factors, such as a need to reduce cost, or an external motivation such as reduced fat or salt in response to a pressure to be healthier.

History is littered with examples of new formulations of favourite products being withdrawn in favour of the original. Most notable among these is New Coke, the reformulation of Coca-Cola introduced in 1985 to replace the original Coca-Cola. Public reaction to the change was less than favourable leading to the subsequent reintroduction of Coke's original formula.

It’s essential to get your research positioning right.If you tell me the product is ‘light’ I may have different reactions or expectations than if you don't. If everyone around you is reducing salt levels in their products you may be able to reduce salt at a faster rate than if you are alone in doing it, as other products will affect people's taste tolerances. But you are likely to be testing products several months ahead of when they hit the shelves so research needs to take this into account.

With a reformulation, we would have identified early on that either this change in recipe would not work or we would have worked out how to position it and communicate it so that it did work. Rather than testing once, making changes suggested by the research and then hoping for the best, we would re-test each iteration of the product qualitatively, so that the optimal version is put forward for quantitative research, which again saves wastage.