What is an “Action Standard”? And why should you have one?
Action standards are especially important in product tests. They determine the criteria, which will decide whether a product will be adopted or rejected; i.e. action standards serve as benchmarks for product performance. As such, they come into play whenever a product is re-formulated or in line-extensions.
Product development is always motivated by business objectives. If one didn’t have a clear objective or goal in mind, why bother developing a new product? As the objective is known only to the client, it is the client who needs to set the action standard; i.e. identify the criteria that will lead to adoption or rejection of a prototype.
Superiority & Parity
Two concepts are important in action standards:
- Superiority: the product must perform significantly better on a number of parameters than either the current version or competitors’ products
- Parity: the product must perform as well as the current of competitors’ products
In both instances, the parameters of product performance need to be well defined; e.g. taste, fragrance, overall acceptability, etc. Again, it is the client who needs to identify those parameters.
Level of significance
Statistical significance tests are used to determine whether a product is rated superior, at par or inferior by consumers. Three levels are commonly reported on by research agencies:
- 90% Confidence level: 10 times out of 100 tests the results may not be significant
- 95% Confidence level: 5 times out of 100 tests the results may not be significant
- 99% Confidence level: only 1 time in 100 tests will results not be significant.
While we report all three levels in our reports to the client, only one level is used to serve as benchmark. The standard benchmark is the 95% level of confidence. Why would a lower level of confidence be adopted if the 99% level seems so much more accurate?
There are two risks in setting benchmarks:
- The benchmark is too low; the product appears like a winner – but goes on to flop in the market place
- The benchmark was set too high; the product is rejected – although in the market place, it might have performed very well; this is a case of missed opportunities…
The 95% level of confidence is neither ridiculously low nor impossibly high; hence it is the most common benchmark
Deciding on Superiority or Parity
As mentioned above, business goals will determine whether superiority or parity are aimed for. A few examples will demonstrate this.
- Assume a scenario, in which a product is being re-formulated in order to lower production costs, thereby increasing margins. The re-formulated product does not need to be better; it just needs to perform as well as the current formulation in order to retain its franchise. Hence, the benchmark will be parity.
- If, on the other hand, a product is re-launched with claims of “bigger and better” or “new and improved”, the development version clearly must perform better than the current formulation; hence, superiority will be postulated.
- A product has been reformulated with the aim of cutting into competitor’s franchise. In this case, we would aim for parity among current franchise and superiority among the competitor’s franchise.
Action standards are a crucial element determining the success of product tests. They must be in line with business goals. No research agency can – or should – second-guess a client’s business objectives. Hence, it is important for the research team on the client’s side to liaise with the marketing and R&D teams in order to formulate meaningful action standards that not only conform with the client’s business strategy, but can then be passed on to the research agency to inform the research process.