Some years ago I drafted four principles to guide the research I do for clients. They’ve stood the test of time.
1. Interesting isn’t enough
My clients need to make decisions.
I start to worry if the only thing people have to say at a presentation of findings is: “That’s really interesting.”
To be genuinely useful, I need to work as hard on understanding the decision-making context as I do on understanding the people the research is with.
2. Big ideas don’t need big words
Research is worthless if it’s not well communicated.
That often means cutting the same research in different ways for different people, using stories and quotations to bring findings to life, or turning presentations into interactive workshops or conference speeches.
It rarely involves jargon, neologism, or tables in 8pt text.
3. Methodology is not a religion
I believe the qualitative and interpretative methods I’m an expert in deliver genuine insight and value.
But I also know they’re not the answer to everything, which is why I love working on mixed methods projects and in interdisciplinary teams.
If the methods I use won’t answer the question being asked, I’ll say so.
4. Fieldwork is the job
The kind of research I do is all about seeing the world how other people see it.
I am not entirely convinced you can do that unless you look them in the eye and hear what they’ve got to say.
That’s why I like to do at least some of the fieldwork on any research project I’m involved in.
Passion, performance, practicality: a segmentation of motorcyclists
In this project, delivered by a multidisciplinary team and using mixed methods, we set out to look beyond the easy stereotypes and engage with the diversity of different people who use powered two-wheelers on the roads.
There’s a link to the full report at the bottom of this page. Here, I’d like to say a few words about the topic of segmentation.
Are there really just seven types of motorcyclist?
Are there really just seven types of motorcyclist? Not six, or eight? Is everyone just one type, or can you be a mix? Can you be one type on weekdays and another at the weekend?
May answers to these questions are: No. Who knows? Not sure. Maybe.
The important point to grasp is that segmentations are NOT descriptions of reality. What a segmentation study like this is saying is: “For certain defined purposes, pretending that there are just these seven types of motorcyclist, in these proportions, will lead you to make better decisions than, for instance, treating all motorcylists as if they were the same.”
The reality, of course, is that every individual is different. Human beings are complex beings, not reducible to a few dimensions or a two-by-two. And human practices are intricate and nuanced cultural entities. There is, we might say, no such thing as a rider, and no such thing as riding: there are a lot of people who engage in activities which happen to have in common the use of a powered two-wheeler.
Decision-making and the value of segmentation
Researchers like me delight in this kind of complexity. A heart, I’m a naturalist: I love nothing more than to wander out into the world human experience and do what I can to record and catalogue its inspiring diversity.
But, unlike me, my clients have to make decisions. How to encourage the use of protective clothing, for example; or how to promote safer behaviours. And from this perspective, diversity is not inspiring: it’s baffling.
Every individual is different: but it is not practical to have a policy, a strategy, a proposition tailored to every individual.
One solution, of course, is to target the average. But in a diverse population, that can do more harm than good. There’s an old joke about a physicist, a chemist and a statistician hunting: the physicist shoots and misses the stag by one metre to the left; the chemist misses by one meter to the right; and the statistician shouts: “We got it!”
Segmentations are a pragmatic response to this challenge. You can’t tailor to every individual. But you can’t pretend everyone is average either. Instead you use the evidence to create a model which will help you get a handle on the complexity.
Segmentations and purpose
A crucial rider (pardon the pun) to all of the above. Segmentations are designed for a particular purpose. They won’t necessarily work for another. Because – remember – they are NOT descriptions of the reality. They are pragmatic and purposive
If you’re manufacturing T-shirts, for example, it makes sense to segment the incredible diversity of human shapes and sizes into, say, S, M, L and XL.
But that segmentation is not going to be so useful if you manufacture off-the-peg suits. And it’s even less helpful if you make shoes.
