A colleague recently remarked: “Why does the market research industry still rely so much on sample averages when marketing today is so targeted?”

A good question. Take innovation as an example – new products are rarely developed for the mass market, but increasingly for small niches where a specific consumer need can be met.

The saying goes, “Safety in numbers”, and maybe that’s why some traditional researchers still rely on averages. But I believe averages hide a lot of valuable insight, so to provide more precise advice for the modern marketing world, you need to look at an individual rather than aggregate level.

This is important for innovation success, so our innovation tools at Kantar TNS use this philosophy throughout. In particular, we use an individual approach to estimate trial, and importantly the growth potential for any innovation (incrementality).

Trial is our headline measure of consumer interest in a perfect world – if everyone is aware of the new product and able to buy it.

To understand trial in the real world we use an individual model based on purchase intent, value for money and people’s attitude to purchasing products in the category – something we call ‘inertia’.

Here’s an example to illustrate the difference in this approach. These two respondents, at first glance, seem to have shown the same interest in a new product based on purchase intent and price perception.

An aggregate model would treat these two people the same, because it applies a correction to the total sample and is therefore only driven by the overall level of purchase intent and value for money.

But, based on her attitude to purchasing in the category, Sandra is less open to innovation and currently has a set category purchasing pattern. Do you believe she will be just as likely to try a new product as Lucy, who is more promiscuous in her category purchasing behavior and more experimental?

We know they don’t have the same potential. Our model assigns a lower probability to Sandra, because a new product will have to overcome her high inertia.

A client once enquired, “Why are older housewives always interested in my products when they are not who I’ve designed them for?” This example explains why.

Older consumers tend to be more brand loyal and less experimental, and hence have high inertia. Exactly the conditions we saw above which led to what we call ‘over-claim’, and therefore the need to downplay their likelihood of trying a new product. So with aggregate analysis, you get a positive but incorrect picture. While an individual approach factors out the over-claim by different subgroups to provide more valuable insight.

And the added benefit of assigning an individual probability is we can use this to weight source of volume analysis. If most of your high probability trialists are existing brand buyers who will switch, the new product will heavily cannibalise your business, and the idea is unlikely to be a strong growth driver.

Through R&D using Kantar Worldpanel data, we’ve shown this more precise application of individual consumer interest to source volume analysis can double the accuracy of our prediction of growth that new products are likely to deliver to your business.

We talked about the importance of incrementality in a previous blog, so given this accuracy, individual modelling provides a significant improvement in our ability to identify and prioritise ideas that have the potential to provide successful growth in-market, early on.

So don’t rely on averages when evaluating the potential of your innovation pipeline. Use individual level data to get a more accurate picture of growth potential, and more actionable insight around launch targeting to optimise the launch success.

Individuals matter.


eValuateNow is a concept testing tool from Kantar TNS that identifies which concepts have potential to succeed in market and drive growth, in as little as 24 hours. To find out more, visit zappistore.com.

David Soulsby

Posted by

David Soulsby

David Soulsby is a Global Director of Innovation and Product Development at Kantar TNS

Loading Disqus Comments ...