How to Build Confidence in Innovative Research

I was in the boardroom with the CEO of our new client, debriefing a project we had just finished. Our contact (the head of insights) was there, along with the heads of product development and marketing, and two of my Irrational Agency colleagues. We had just delivered a project based on a newly designed behavioral-science-based methodology, and I was demonstrating what consumers wanted from the company’s new product range.

Then came one of the worst moments I have ever had in a research debrief. The CEO turned to me and declared: “I just can’t trust these results. I don’t believe what you’re saying.” I felt a dark knot forming in the core of my stomach. A flood of thoughts rushed through my head: had I made our contact look bad? Had we wasted the client’s money, and our time? We certainly weren’t about to win any more business from them.

Of course, there are always client stakeholders who don’t really want to listen to research, and who are convinced they know best. But this felt different. This was a failure on our part to demonstrate why our methodology was accurate and robust; a failure to communicate the findings persuasively; a failure to build confidence and faith in the process and in our agency. And it told me we had to find a new path.

Presenting innovative research requires extra care

“[W]hen a client is trying out a new method for the first time…the methodology is on trial as much as the results.”

This is a peril for every innovative research project. Maybe it’s not so bad if you’re repeating the same tracking study or the same customer satisfaction survey from last quarter. The default assumption is that the data is correct, and any unexpected results are showing a genuine change in customer attitudes rather than a research mistake. But when a client is trying out a new method for the first time (as they often are, when it’s narrative research or a behavioural science method), the methodology is on trial as much as the results.

And the risk isn’t only to my hurt feelings. If stakeholders don’t trust the research, they won’t act on it. The research budget will be wasted and the business will make worse decisions without guidance.

Seven tips for successful innovative research

Since that depressing day, my colleagues and I have developed a new approach to building confidence in the research results from innovative methodologies. Here’s what we do now:

1. Gain a robust understanding of research stakeholders

There is simply no substitute for that intuitive understanding of what clients need to know, the language they speak, and their expectations. That intuition? It grows with experience. Now, our team is led by senior researchers with 20+ years of tenure talking with stakeholders.

2. Use a repeatable, solid set of tools

Instead of inventing a bespoke methodology for every project, we now put together a project from a few jigsaw pieces: the well-tested tools we have used with dozens of clients before. With years of experience in each tool, we know how to best use them and how to answer the questions that clients ask. The approach is still innovative, but it’s not a bleeding-edge experiment every time we meet a new client.

3. Conduct public research results to share in advance

Client confidentiality used to make it difficult to share examples of past research, to show the reliability of an approach. But we now publish a quarterly trends report, allowing stakeholders to see a detailed example of what they will get before commissioning a project. Stakeholders rely on the insights from this report to guide their business, even if they haven’t yet carried out ad hoc research.

4. Advanced science – but in the background, not the foreground

We base everything we do on science that is validated by hundreds of researchers worldwide, across psychology, economics, and neuroscience. But that doesn’t sit up front. The business question is the priority: the science is only there to demonstrate the credibility and reliability of the methods used to answer it.

5. Balance the expected and the unexpected

Everyone wants to discover something new in a research project – otherwise why bother? But if everything is new, it is hard for them to connect the insights to what they already know, and they may not believe what they hear. After all, how likely is it that one project would overturn all the accumulated knowledge of a brand team about their category and product?

6. Show that behavioural science methods are more trustworthy than traditional research

Conventional research has some well-known flaws, such as consistent overclaiming of purchase likelihood, and respondents’ tendency to agree with conventional wisdom. By showing how behavioural science makes these answers more accurate, we demonstrate how this research is more reliable than the old approach.

7. Apply data and consistency checks

If something does go wrong in a research method, there is usually some kind of early warning. If the data doesn’t make sense, you can spot this either in the individual respondents’ behavior, or with statistical checks. We have introduced a checking process in our qualitative research, and at three different stages of our quantitative surveys: so there’s always time to notice if something is wrong. In most cases everything is fine, but when you’re innovating, you need to be alert to this possibility and know how to correct it.