If a product test is being conducted to assess a replacement recipe rather than a brand new recipe / SKU, we always ask the client if their aim is:
· for the replacement recipe to score at least as well as the existing recipe. This is often the case when the new recipe has been constructed to reduce costs – slightly less chocolaty, if you will
· or for the replacement recipe to outperform the existing recipe. This is usually the case when the existing recipe has disappointed. The reformulation might make it ‘slightly more chocolaty’ or up the fruit quotient, for example.
Once we know the answer to this, we set our parameters for the analysis of the quantitative data. In the first instance, it’s a straightforward parity test; if the existing recipe gets an average 75% taste approval, then so must the new recipe. More critically, when the two products are put head-to-head, the replacement only needs to have the same level of preference as its predecessor.
The second option – looking for improvement – presents a slightly different challenge. Here we advise the client opts for a 60:40 test. Simply put, this test stipulates that, head-to-head, the new recipe must get a minimum 60% of the votes. This is our benchmark; at the 60% level, it is evident that the new recipe is preferred by a substantially higher proportion of the consumer base.
It’s a very simple yet powerful technique.