Mark Ritson: Those who say demographic targeting is dead are wrong, it’s still a valuable activity

It seems everywhere you look these days there is a marketer proclaiming the death of demographics. From experts at Mindshare to JD Power and from publications as diverse as CMO.com and Brand Quarterly, it would appear that the era of demographic targeting has come to a close.

Unsurprisingly, much of the criticism has come from the US. In a year in which Caitlyn Jenner has highlighted the transgender issue and Rachel Dolezal has done the same for the transracial discussion, the continued use of demographics to portray and predict a consumer’s references was always going to be contentious.

On a superficial level these critics have a point. Targeting the 28- to 40-year-old mum is clearly not the optimum way to market any product or service. But that was always the case. There have always been shit marketers who don’t do any research or segmentation and just conjure up a broad, stereotypical ‘target segment’ to make their marketing plan look more professional. In that sense, demographics have always been dead. A point I would encourage those obsessed with so-called ‘millennials’ to thoroughly contemplate.

To be fair, most decent marketers never used demographics this way. They recruited a representative sample of the market and asked for demographic questions along with a bunch of attitudinal and behavioural questions. Then they did behavioural segmentation in which the market was sorted into distinct sub-groups based on how they thought and what they bought. Only then, with a small distinct segment of consumers identified, did we apply the demographics and examine if a segment was demonstratively more likely to be older, female, urban etc. There’s a huge difference between assuming girls like pink and boys prefer blue and a representative survey of the population showing a statistically significant skew in colour choice between male and female respondents. The difference between stereotypes and segments is data.

My defence of demographics may be outdated, however. The entire approach to marketing in which you commission research and segment the market may be going the way of the dodo if experts are to be believed. In the era of digital marketing, the premise of needing to segment and target in the traditional manner is transformed. Once you have a mountain of big data attached to a specific consumer derived from past search activity or from Facebook interactions, the relevance of a consumer’s gender or age becomes relegated to an afterthought. Despite what the financial services industry might warn, past performance of consumer behaviour really does predict future performance. Marketers can buy eyeballs based on what they have previously looked at, rather than the age or gender of the skull that they happen to reside within.

But hold on again. This idealistic vision of purely behavioural segmentation devoid of any and all demographic identification depends on a couple of assumptions about the market. First, that you have this data for close to 100% of the target audience and that you can reach them all using media that allows you to identify each consumer individually. That may appear to be possible for the addled social media obsessives who now seem to dominate marketing and assume every campaign is a blend of Pinterest, Twitter and Facebook.

But only half of the UK’s marketing money is spent on digital communication, the other 50% is invested in ancient media monoliths such as TV, radio and print. For those methods –  and therefore half the marketing being done in the UK – we  need some form of demographic data. Even in the brave new world of programmatic don’t forget that much of the data being modeled inside the black box is good old-fashioned demographics.

Brute demographics without other data were always stupid. But using demographics as part of a broader approach to behavioural segmentation and targeting remains a valuable activity for many marketers and brands. If you have complete behavioural data and the ability to buy targets based on past purchase and current interests, go for it. But don’t throw the baby (or any other age-based segment of the market) out with bathwater just yet.