Rory Sutherland: Taking aim

x Sponsored content on Research Live and in Impact magazine is editorially independent.
Find out more about advertising and sponsorship.

The very phrase ‘target market’ is revealing. It brings to mind a set of concentric circles with a bullseye in the middle. This bullseye is of course your ‘core’ target market or audience.

Like ‘audience’, ‘market’ is used in the singular. The driving idea being that the closer someone is to some demographic or notional ideal, the more attention and expense should be devoted to their conversion. And that, just as a target has only one centre, a target market has at its centre one ideal customer with other, less valuable customers radiating outwards, progressively diminishing in importance in proportion to the extent they differ from that ideal in one or more characteristics.

I suspect this idea is popular in some digital circles, since a linear model of that kind is amenable to optimisation algorithms and so forth. Whether it bears much relation to reality is a separate question.

Recently, a number of prominent campaigns have diverged very significantly from the idea of the ‘target market’. These approaches have been characterised by ‘splitting not lumping’. They also test a far wider range of messages and creative approaches than is typically found in online advertising. Notably, these new approaches have been used not in commercial marketing but in either political campaigning, or in combating vaccine hesitancy during the pandemic.

They have all diverged from the textbook approach for fairly sensible reasons. In the case of promoting vaccination against Covid-19, your target audience is, of necessity, everybody. Many of these people will be readily compliant, but even these easier targets will need some encouragement or helpful reminders to overcome inertia. At the other extreme, some of your audience will be highly sceptical or downright hostile from the outset.

In the case of the resisters, their reluctance will have many different causes and be driven by very different motivations and beliefs. In such cases, different messages may be necessary to overcome their existing antipathy – for instance an appeal to the collective good may work better for some; an appeal to fear may work better for others. (There was even a group identified whose anti-vax stance was predominantly driven not by any ideology, but by a fear of needles.)

What is noticeable about this work is that it reversed the typical marketing approach – ie identifying people and matching a message to them. On the contrary, it identified messages and then set out to find those people with whom the messages resonated most.

In the case of political campaigning, it is acknowledged that some people are impossible to convert, others will vote for you regardless. Nevertheless, in pursuing the magic margin of floating voters, it was accepted that different messages would exert different levels of potency. There are some political issues, after all, which many people just don’t care about at all. Equally, many people have a particular hot button. Ahead of the Brexit vote, alongside money, migration and one or two other emotive issues, the Leave campaign at one time considered creating a group called Vapers for Brexit to mobilise a group of a million or so people who felt threatened by the EU’s highly restrictive approach to e-cigarettes.

I would call this a ‘behaviour-first’ approach. Rather than identifying the perfect message for a theoretically ideal customer, you identify those diverse pockets that might be amenable to adopting a behaviour and work backwards from there.

In consumer marketing, this approach is applicable more often than we think. I hope Byron Sharp (professor of marketing science at the University of South Australia) won’t mind me saying this, but one of the most important messages from his research is essentially that ‘your customers are a much messier group than you like to admit’. The idea that brands are distinguished from one another by demography is largely baloney.

There are several other reasons why I think the ‘splitter’ approach may be underused. For one thing, digital and addressable media are perfectly suited to it, in a way that mass media are not. For another, many large advertisers – insurance comparison websites, broadband and satellite TV, mobile phone networks and so on – effectively have a target audience of ‘everybody’.

But there is a bigger reason, still: the value of serendipity. By starting with multiple hypotheses rather than a single preconception, you are simply much more likely to get lucky. True, you sacrifice a certain theoretical and procedural neatness, but the great gain is that the second approach is much more likely to reveal something valuable that you didn’t know, rather than merely confirming (or disproving) what you think you know already. Define your audience too narrowly and you may end up with a self-fulfilling prophecy.

This column was first published in the July 2022 issue of Impact.