• Idea Lounge
  • A New Behavioral Science Market Research Tool: The Attribute Elicitation Task
A New Behavioral Science Market Research Tool: The Attribute Elicitation Task
View transcript
When you don't know what you don't know.

Regardless of what sector you’re working in, what product or service you're offering, at some point in developing your marketing strategy you're going to have to figure out what is it that your end-user is valuing about your offering. That brings up the question, how do you go about figuring that out, figuring out what you don't know when you don't know it? What is the scope that that can encompass? This is a challenge that's faced by almost all our clients in marketing, at some point along figuring out their marketing strategy.

Currently, I think, the options that clients have is this trade-off between going the quantitative rout or the qualitative rout. When you go the quantitative rout, the benefit is it’s scalable, so this means you can talk to a lot of consumers, across segments and across markets. But, the challenge with the traditional quantitative approach is that the structure means that as the marketers you have to decide what you think might be important to your consumer. You put that in front of them, but you can only get feedback on those specific factors. The risk here becomes if you miss the attributes that are actually drawing from the behaviour your miss-spending your marketing dollars. The other option is to go the qualitative rout. The benefit, of course, of qualitative is it’s a little bit more open ended, so there’s not that large risk of missing the most important attribute. But we still have to remember that we have interviewers who are conducting that qualitative research and their own biases and expectations are still going to direct the conversation and direct the insights. Assuming you have perfectly objective moderators (and that’s not going to happen) the other challenge is that, when you’re doing qualitative, it’s not quite scalable. The risk here becomes you have to extrapolate the really small sample size to larger samples sizes and, again, this is risk way to spend your marketing dollars.

I think the current innovation in this space is scalable qual because it has the scale and it has the open-endedness. But I don’t think the problem is quite resolved yet, and the reason for that is, whether we’re talking about the scalable qual, whether we’re talking about traditional qual or traditional quant, mostly these are resting on explicit measures. Directly asking people “what is it that you value?” “what’s important to you about this offering?” And what we know from Behavioral Science is, and am I’m sure you’ve heard a lot about it today or will today at this conference, it’s not very easy for people to 1) access what’s actually important to them or 2) actually communicate that to us.

What we’ve developed is the Attribute Elicitation Task. This task is scalable, so we don’t have to risk extrapolating from small to large markets. It’s structured in terms of the task, which means we don’t have the risk of the interviewer directing or biasing the research, but it’s open-ended in terms of the responses we get back. And, what that means, is that the respondents are able to tell us what’s important to them rather than doing that the other way around. And, finally, we use an implicit measure that gets over the traditional rationalizations we get when we’re using direct line inquiry.

I’ll very quickly take you through what this method looks like. It’s operationalized in four stages. In the first all we do is present them with three products and we’re just asking them to decide which two of these are similar and what makes them similar and how is the third one being differentiated from the other two. As you can see, there is structure here and there’s no value judgement. We’re not asking them “what’s most important to you?” “what’s driving your decision making?” we’re going to get that from this implicit measure of just seeing how they’re naturally categorizing.

You can see here that it is completely open ended. There’s a tone of different ways that they can put these together, based on what attributes are important to them. Once we figure out what that attribute set is, by giving them a bunch of these triplets, so those three quadrants we call triplets, we give them sets of those to illicit a set of attributes. We then ask them to explicitly tell us which one of those are important, but we pair that with an implicit measure of importance, and that we get by just analyzing how often did an attribute come up and how quickly did it come up.

Once we know what’s important to them, we then get them to benchmark how’s your product fairing on these attributes compared with competitors and compared with an ideal. This allows us to say how are we performing, how are you performing relative to competitors and often times also highlights where the whitespace is. Once consumers are completed with this task the output that we get is this grid. We essentially have a list of attributes that are important, relative to your product, and we know how you’re delivering or not on those attributes. Once we’re able to plot importance versus what the attributes are we can give you a really principled understanding of where to invest your marketing dollars.

We have successfully applied this with a bunch of clients in different industries, and again I think the reason we’re able to apply it so broadly is because at some point, it doesn’t matter what sector, or what the product is, you’re going to be faced with this question of figuring out what is the value proposition of your portfolio.