back to behave arrow

Humbug, frameworks and 'what kinda?' questions

Mark Earls
Independent marketing and advertising professional
Copy Copy Copy

Mark Earls wonders, will asking simpler questions be more useful in understanding behaviour?

If you haven’t yet got yourself a copy of Paul Feldwick’s excellent new book, An Anatomy of Humbug, I heartily recommend you do. It is – by a country mile – the best book I’ve read about how we think about advertising and how it works (or how we’d like to think it does). Typically eloquent, insightful and generous in equal measure – a must-read for all those who create, sell or buy advertising, and for all those researching advertising. While Paul readily admits that his own preferred model increasingly tends towards what he calls ‘the Barnum one’ (the ‘humbug’ of the title), he’s characteristically generous in accepting that many of the models have something useful to teach practitioners. Like the partial perspectives that the blindfolded mahouts reveal as they touch different parts of an elephant, each of these models reveal something about the whole – not the whole pie but something, at least.

More than one model

This is not how the conversation about advertising and advertising research tends to run. Much of the time, advertisers and researchers assume that there’s only one way that ads work and only one superior way to research it. It’s a monomaniac research arms race.

There are exceptions of course. When Hall & Partners’ own framework approach was introduced more than 20 years ago, the focus of the industry’s response was definitely driven by the inclusion of the more emotional and salience-based models (the kind that creative agencies instinctively work from), alongside the more rational and persuasion-based models that reflected how (particularly US-based) clients tended to think. Nowadays though, the striking characteristic of the Engager® framework was that it acknowledged publicly that things are not singular. There have always been a number of different ways in which practitioners, vendors and buyers think about how advertising works, not just one.

Accepting the many

For my money, the real virtue of accepting this is that it forces us to step back and ask ‘what kind of’ (what kinda) questions before rushing off to measure anything: what kinda thing is this, and what kinda measures are appropriate to use? The measures that are appropriate to salience, for example, aren’t going to be what you need to measure persuasion-type advertising.

Too often we treat the bigger marketing and behaviour change challenges we face as if they were singular

But what kinda questions have a much broader application than advertising and advertising research.

Too often we treat the bigger marketing and behaviour change challenges we face as if they were singular – as if no one had ever faced such a thing before (which is, of course, very flattering to the person who gets to solve it, and to the heroic client who buys such a solution).


Asking what kinda questions helps you access the right kind of learning from elsewhere, whatever you’re trying to do.

Here’s a striking example involving what kinda questions. A decade ago Professor Martin Elliott – one of the world’s leading pediatric cardiac surgeons – and the team at the world-famous Great Ormond Street Hospital in London were searching for a solution to improve patient outcomes at a critical moment in their surgery; the moment when, after ten to twelve hours in intensive and stressful theatre, the tiny patient gets transferred to ICU and a whole new team. Previous research had identified this moment as a crucial one for human-factor errors – simple and understandable mistakes made by a large team of specialists under conditions of extreme fatigue, stress and technical difficulty.

Slumping on a sofa to watch the Grand Prix highlights after a long day in theatre, Martin was struck by the similarities between the real-time challenge that the F1 pit teams faced in the high-octane race world, and that of his own team: many specialists, extreme high-tech equipment of different sorts, huge time pressure and life-critical outcomes for the vulnerable human at the heart of the process. So working with McLaren and Ferrari, Martin and the team have managed to establish a new protocol for themselves, ported directly from F1, and in doing so reduced human errors by 42% and patient outcomes by a similar margin.

What kinda questions and strategy

To turn this approach into a useful strategy tool for marketers, I’ve made use of the data-driven model of choice styles described in a previous book, I’ll Have What She’s Having, written with Professors Alex Bentley and Mike O’Brien. What kinda questions are job #1: how are people in your market choosing?

Then, to help you find the right kind of things to copy, we’ve sorted successful strategies from a huge range of different contexts into each of the four kinda boxes. From mainstream brand marketing to 18th-century nutritional campaigns, from leading-edge experiments in public policy to the immortal Blues Brothers and their ‘one night only’ promise. Once you know what kinda behaviour you’re trying to change, it becomes a whole lot easier to find inspiration and things to copy in unusual places.

Maybe there’s also a simple learning for market research practitioners in general; before we rush off to find new and better ways to measure phenomena more precisely – using whatever fancy tech or scientific techniques we have at hand – perhaps asking a simpler kinda question might be more useful.


We know now what we do

We know not what we do


Every breath you take


Creating meaningful conversations