Jigsaw is the data-sharing consortium created as a joint venture between Cadbury Trebor Bassett, Kimberly-Clark and Unilever. Agency OgilvyOne has been looking at ways to exploit the full potential of the database and improve the quality of the services that Jigsaw offers.
The key is to achieve a balance between acquired, modelled and enhanced data.
It's no wonder that this is a common dilemma among direct marketers.
Actual data collected at an individual or household level is more accurate than modelled data, but is - by definition - much more expensive to acquire.
Modelled data offers a multitude of variables and targeting possibilities, but does it damage your brand to make assumptions about customer behaviour and demographics? Or is it better just to buy cold data on rental basis?
Confusing, isn't it?
The solution, of course, is to have a considered (and thoroughly tested) balance of actual, modelled and cold data. But where to start?
Look closer
The first step is to understand your existing customer data at a very fundamental level. A full data audit will tell you the number and population level of the variables you already have. More importantly, it will reveal the gaps that you need to fill - either through acquisition or modelling.
You should also look at the level of usage surrounding these variables.
For example, why spend money acquiring and updating a variable that has never been used for marketing or insight activity?
The audit should also reveal what your key customer variables are - that is to say what are the core attributes that define a good, bad, or indifferent customer. These will usually be few in number, and will generally be a mixture of lifestyle and transactional criteria, such as household composition, date of birth, income, and purchase history.
Approach a specialist database bureau (not necessarily your database host) to undertake the audit for you, and always ask for sample audits in advance to make sure that their reporting has a satisfactory level of detail.
Lay a sound foundation
Understanding your key variables is invaluable in guiding your acquisition activity. Concentrate your budget initially on the criteria and golden questions that will allow you to segment and contact your customers by lifestyle, lifestage, and household type at a base level. Name, address, age, household composition and income are the foundations - the actual data - upon which you can build, extend, and refine through modelling, enhancement and profiling.
Seek out more data
Collection of these key variables can be done in a variety of ways, but it is most cost-effective to use existing consumer touchpoints - such as web, careline or promotional activity - to collect a few additional criteria above and beyond the usual name and address details. Consumer touchpoints are where you can find out about consumer attitudes towards your brand. And get permission-based (opt-in) e-mail addresses and mobile phone numbers too for multi-channel communication. Even better, find out what your customer's preferred channel is.
Then overlay modelling
For some brands, key variables alone will provide almost all of the targeting and insight criteria that they require. In quite a few cases, household composition, age of children and income are far greater indicators of consumer type and propensity than more complex lifestyle criteria (for example, household composition could allow fairly accurate prediction of washing powder consumption without knowing actual or claimed purchase behaviour). Once this base information is present, however, lifestyle overlays and catchment area mapping, among others, can be used to develop and refine models to even greater accuracy. Modelling takes on a whole new dimension when applied as an overlay to - rather than an as a substitute for - actual data. There are myriad tools, data providers and analytical techniques which can transform your data from being merely functional to a highly predictive asset in this way. And get competitive: your highest response or ROI result is merely the benchmark you have to beat next time.
Your data journey should be a constant learning experience.
Mind the gap
In certain cases though, neither core attributes or modelling will provide the specific criteria needed to target specialist markets or consumer-types. In these cases, in-fill can be achieved through tactical list-buying, or survey question sponsorship. The rule with any cold list activity is test, test and test again. Don't commit or roll out until you know it's going to work.
Effective data strategy then is about understanding your existing variables and your gaps. Armed with this information, you should be able to apply your budget where the most targeting and insight benefit will be gained.
CASE STUDY: JIGSAW
OgilvyOne has been helping many of the leading FMCG brands within the Jigsaw consortium to collect core household and claimed-purchase information (around 200 variables) for millions of UK households via offline and online survey activity.
Analytical tools
The agency has a dedicated team providing offline and online campaign-generation, coupon-redemption analysis, predictive modelling, post-campaign evaluation and catchment area modelling, using a variety of analytical tools, techniques, and brand and CRM experience.
A key goal is understanding customer value to a brand, and how this information can be used for differential marketing, ie understanding what the marketing spend for an individual should be, based upon their short and long-term value.
This is primarily achieved through predictive modelling based on household composition and a combination of claimed, actual and inferred purchase data. Another important area is couponing, and how redemption, or lack of it, affects consumer value and purchase behaviour.
One issue for FMCG companies is that actual purchase data is often difficult to obtain in representative volumes. Therefore, propensity and predictive modelling becomes crucial in helping brands to segment and target effectively, and makes the substantial volumes of claimed-purchase data available within the consortium highly valuable.
This information is used to develop effective relationship marketing programmes, with a particular focus on consumer value and behavioural insight.