Direct route to primary target

The DM industry has developed a vast array of tools to target consumers, but whick work best? David Reed learns of a way to bridge the gap between individual data sets and geodemographics

There is an equation which explains how direct marketing works: budget ÷ population = targeting. Since you cannot afford to appeal to everybody in the country, you have to find ways to narrow the field.

Over the decades, the DM industry has developed increasingly accurate ways to target and direct marketers have an array of alternatives to consider. From geodemographic classifications of the entire population to near real-time hot leads from current purchasers, there is a system to suit almost any pocket.

However, many in DM agencies are disparaging about what is on offer. “Demographic classification systems are solutions looking for problems – they are the lazy weapon of choice in the data planner’s tool kit,” says Neil Johnson, data planning director at Claydon Heeley.

Jed Mooney, managing director of Datahold, believes the reasons for dissatisfaction with prospect data lie in improved access. “Analytical tools are becoming more accessible via software on the marketer’s desktop. So the way direct marketers are selecting data is becoming more intelligent,” he says.

DLG communications director Richard Webster adds that geodemographic systems are not targeted enough for direct marketers looking to employ direct mail. “The weaknesses of geodemographics are particularly evident in urban and cosmopolitan areas, which are typically where the young, affluent targets for marketers live,” he says. “Very different groups are living close together, which mixes up the assumptions made 25 years ago.”

Against this he sets individual-level data, directly derived from the consumer, which allows for the personalised offers and targeting the DM industry advocates. For some advertisers, the single most valuable variable they are looking for is that an individual is currently looking to make a purchase. Hot lead data of this sort is not cheap, nor plentiful, but it is powerful.

“Consumers have freedom of information, movement and economy. So it is harder to make a judgement based on where they are living or limited demographics. We live in a more flexible society than ever and want to be treated as individuals,” he says.

This is the most powerful argument for the new generation of individualised data sets, but geodemographics are not going away any time soon, either as a core marketing tool or as influences over each and every one of us.

Webber can justifiably be called the godfather of geodemographics, having developed the first-ever classification system, Acorn, and its major rival, Mosaic. He has just completed work on the UK’s first ethnological classification, Origins.

“The power of geography is more true than it has ever been, but it is very easy to misunderstand,” says Webber, now managing director of Origins Information.

Neighbourhood knowledge
A major criticism of geodemographic clustering is that diversity in each segment is considerable. But Webber points out: “People in one type of neighbourhood are no more diverse than people of the same age or set of qualifications. Knowing occupation is less predictive than knowing neighbourhood.”

Difference is the core of what marketers are looking for – any indicator that their specific target audience can be identified from among the mass. They also want some prediction that their campaigns will influence behaviour in the right way. Knowing that a geographically-defined cluster is more likely to respond is a powerful way of getting this uplift.

Hybrid solution
A hybrid solution that seeks to bridge geodemographics and individual-level systems is Personicx from Acxiom. Based on directly-derived data from 60% of the UK population, it uses 3,000 variables to generate clusters that are highly predictive of buying behaviour.

“Personicx is behaviourally optimised,” says Ian Stewart, head of product sales at Acxiom. “It maximises understanding at an individual level, rather than the segment level in geodemographics.”

Stewart adds: “Many clients have built predictive models for crossand up-selling to existing customers. But when it comes to prospecting, they have no way to access and action those same models.”

Paradoxically, the way those models can be overlaid into Personicx is via the geodemographic system Acorn. Acxiom has those codes associated against its own clusters, so if a client is using them to identify its most valuable customers, it can pick out the same people in the prospect pool.

Choice is one of the hot words used by marketers. When it comes to the tools available to themselves, they are now spoiled for choice, and the reality is that, unless you are mass market, you need to segment. The only way to work out which system is best is to test it.