MRG Evening Meeting: Marketing Mix Optimisation

Last night’s MRG evening meeting featured a presentation from Martin Haywood from Dunnhumby on their Marketing Mix Optimisation (MMO) tool that has been created from Tesco clubcard data.

Tesco is the UK’s largest retailer, with an incredible £1 in every £8 in the UK spent in store. Its loyalty card, the Tesco clubcard, now has 13 million members, which together create 4.8 billion pieces of information each week. Ten years ago there wouldn’t have been computers equipped to even deal with this amount of information. Dunnhumby works with this clubcard data to look at shopping patterns and behaviours.

Martin Haywood believes that the current media industry bases too many of its decisions on estimates, and that sample sizes are too small to be able to make correct decisions. The industry does not do enough with price promotions, image, competitive activity, distribution, seasonality and brand heritage – all of which have been proved to affect sales.

Dunnhumby states that through their MMO they look at what people do, why they do it, put it in its correct context and then take a look at the future.

Haywood stated that the interaction between brand and sales expenditure is rarely fully explored and the line is beginning to blur more and more, with clients increasingly looking at these effects, especially over the next year, he predicts.

Tesco research has shown that only 10.5% of brands have a greater than 5% household penetration in the UK, while 79.5% have a household penetration of less than 2%, demonstrating the long tail of the FMCG industry. Most brands are not used by most people, so the days of when 1,000 people was a robust sample have gone as there is too much choice in the market. Therefore a larger sample is needed, which Haywood believes the Tesco clubcard data provides, with its millions of members.

There is no such thing as an average customer, so the larger the sample the better, according to Haywood. Plus with a larger sample it can be more effectively sliced into target markets, whilst still providing robust numbers.

Haywood believes that if you’re going to work in marketing in the next few years, you have to admit that you will be working in direct marketing.

He believes greater response rates can be achieved through using this data, with effective targeting of offers at the till point. For example, the simple till coupon that follows the receipt has a response rate of 20%, whereas normal direct marketing response rates are closer to 1-2%.

Dunnhumby: www.dunnhumby.com