Quantcast
Channel: Science Rockstars » Ron Voorn
Viewing all articles
Browse latest Browse all 2

What if it was illegal to ask people questions?

0
0

That was one of the most intriguing questions I heard in 2013. Who better to ask than the man who wrote a paper on this, Tom Ewing, Digital Culture Officer at BrainJuicer, one of the most sexy market research companies around.

Tom, what do you do exactly at BrainJuicer?

Well, I work for the Labs team where my role is to help keep track of all things related to decision science, psychology and our own projects and synthesize that into information, insights and methods that can give our customers opportunities to improve their business. And I write a lot.

Tom, you have a background as an Internet analyst, social media researcher and journalist and you won the ESOMAR Excellence Award for the best market research paper of the year for your paper “Research In A World Without Questions”, what was it all about?

It was meant as a kind of compendium of new research methods, and its starting point – from an idea by BrainJuicer founder John Kearon – was suppose it became illegal to ask questions. How would we find out what people think?

In the paper I talked about how unreliable answers in market research can be and what the industry could do about that. They are mostly post-rationalizations of decisions and often have nothing to do with how we really make our choices. Daniel Kahneman’s insights and his thinking on system one and system two thinking have really driven acceptance of these ideas. At BrainJuicer we very much feel market research has been too reliant on methods which force people into “system two” we want to tilt the balance back in a “system one” direction.

How should this then be resolved according to you?

We approach it by inventing methods which get at more system one responses, but also by creating interventions to change behaviour and testing those in the real world.

So you are also into Big Data then Tom, because that seems to be the magic word nowadays?

Big data is exciting because it’s real world, behavioural data. But to be meaningful you need to be approaching it with the correct understanding in place of why these decisions and behaviours happen.

Can you give us an example of one of your more unconventional methods Tom?

A couple of years back we developed “DigiViduals”. It is a technique by which we create digital personas on social media to create insights into what real people do, where they go, what they like, what they eat, what their hobbies are etc.

Please tell us more?

Think about the pen-portraits you do for segmentations. Then think “what if that pen-portrait could Tweet, or go on YouTube, or otherwise social media. What would it do?” That’s one use of a DigiVidual. These fake personas are basically algorithms – they trawl social media and post things that ‘interest’ them, in other words what their algorithm tells them people like them online are interested in. That includes images, music, even location data via Google Maps. Each DigiVidual creates a raw feed of this information.

It certainly sounds very intriguing Tom but how does this help?

When we have all the raw data from the live feed than our qualitative researchers analyse it and give you a much fuller, more insightful picture of how these target audiences live. Consumer understanding tends to usually be through the lens of the brand – DigiViduals gives a fuller picture of the life the brand fits into. Which helps with insight, and helps creatives to develop better advertising.

I also saw a cool example in your paper on how stressing system one gave one of your clients  much better packaging. What was that all about Tom?

One of our clients had a problem with their packaging. They had very information rich packaging which actually tested extremely well. They almost always won in the fake store test situation. But in the real stores their competitors outsold them. If you think about the way people make decisions in supermarkets, it’s very system one – people basically do not think when they are in a supermarket. They mostly want to get out as quickly as they can. We decided to change the way we tested packaging and created an experiment with four groups (see Figure 1).

systemonepack

Figure 1. Experiment on packaging with four groups where Brand A is the competitor.

The first group was not given any special instructions and had as long as they liked to choose a brand. The second group was put under time pressure, the third group received a cognitive distraction exercise and the fourth group got both the time pressure and the distraction – all ways to prevent system two from overriding people’s fast system one decisions. And lo and behold we saw that when we put system two under pressure, the system one reaction was to buy the competitor. Which is what also happened under real conditions. Why? The competitor had a much bigger picture and far less information on the pack and that made it more emotionally appealing, which helped guide the fast, system one decision.

How do you think research agencies will evolve Tom? Do we still need them in the future?

A lot of interesting stuff is coming out now from people focusing on behaviour change – the UK Government’s Nudge Unit, for instance, which is going to be privatised. At BrainJuicer we’ve got an internal unit called the Behaviour Change Consultancy which works at a high level in client organisations to push through understanding of how decisions work – and the more people understand the decision-making mind, the more research will change. The other development is behavioural data – a lot of which can be collected automatically. Mobiles and location data for instance gives so much information on shopping habits and other behaviours.

I think we will see more and more work in developing new techniques on the basis of behavioral insights, and more agility in research. When researchers talk agility they generally mean in terms of response to clients, but what about proactive agility? You have all this data and you can act very quickly – so why not help by experimenting, and by developing and proving hypotheses?

For instance in advertising research, you see it becoming far easier to edit and tinker with test ads – a bit like A/B testing. We can come up with a hypothesis – like voiceovers holding an ad back emotionally, for instance – and rather than just present it to a client we can create an alternative ad and test it quickly ourselves. That kind of proactive agility seems like the right response to change and adept to a data-rich environment.

Thanks for this interview Tom and please keep us posted on any new techniques being developed!

The post What if it was illegal to ask people questions? appeared first on Science Rockstars.


Viewing all articles
Browse latest Browse all 2

Latest Images

Trending Articles





Latest Images