Get in your box: Assumptions in research

I know – what I’m about to say will probably go against everything you know, and believe in, working in market research. But hear me out.

How many times, in the data and insight industry, are you looking at results by a certain sub-group? Whether it be gender, age, socioeconomic group, location, ethnicity, segment, or a multitude of other things.

I get it, I do. It’s helpful to know that your advert resonated more with over-65s than with those aged under 25. It’s important to be aware that your brand is perceived in a completely different way by those in London, and those in the north east. It’s critical to understand how the drivers of purchase differ based on your household composition.

In the research industry, we’re often in the business of putting people in boxes. I am a white, female, 30 to 35-years-old, Welsh, lesbian, parent of a young child. Those are my boxes. Those are the quotas I fill.

The bit I think we forget sometimes is that I’m also an individual. I’m Bethan. I’m a chocolate addict, a passionate Welsh speaker, a stats geek, a cuddle-loving, shoe-hating, real-life person who is just rubbish at spotting sarcasm.

The problem with putting people in boxes all the time is that you lose those nuances, and that colour, often replacing them with generalisations and assumptions, such as “old people aren’t tech savvy, women do the grocery shopping, and Northerners are friendly”. These sweeping statements may be true the majority of the time, or that group might be significantly more likely to display those characteristics, but it won’t be true for everyone.

We recently did some training on unconscious bias, and the danger depending on assumptions and generalisations like these can bring. But what do we do when our industry depends so much on these generalisations?

In my mind, it’s a case of being aware of them, and using them, but not depending on them or defaulting to them. For example, we recently completed some work about make-up and skincare. The original brief mentioned how our sample should be completely female, because, you know, we were talking about make-up. I instantly messaged a close male friend of mine. Turns out, I had good reason to – he owns make-up, and I don’t.

Instead of starting with the assumption that we should speak to females, we started with other qualifying questions to ensure that our sample (consisting of any gender) was in the market for these types of products.

In some instances, this may not make a huge difference. But it could highlight that there’s a huge market for make-up and skincare for males, for example. You could be missing out on a huge opportunity, or a useful insight, because you started with assumptions, instead of facts.

I often quote my Nana in my online ramblings, and today is no different. She has always warned me: “Bethan, when you assume, you make an ass out of you and me.” Let’s not assume, because that is not a box that anyone wants to be put into. 

Bethan Blakeley is analytics director at Boxclever