I Can’t Like a Spec
A car has to be more than a sum of its parts.....right?
Several years ago, my husband was trying to decide which car to buy. He created an extensive spreadsheet with lots of metrics and scores. He spent hours researching and scoring. When I reviewed his spreadsheet, I was baffled. I asked him ‘Where’s the column for how much you like it?’
In his mind, the best car would have the highest score. In my mind, nope, that doesn’t add up. It doesn’t matter how much you weigh the scores from all the specs, I did not believe that ‘liking’ could be just a combination of scores from specs. You often can’t account for just liking something and feeling that it’s the right choice.
To me, this is the difference between qual and quant research. While we want to neatly use qual to explain quant (and vice versa), it doesn’t always add up.
What happened? He tacitly agreed that something was missing, so he ended up making a column for ‘style points’. He ended up buying a car, but I don’t think it was strictly based on specs.
This story represents the difference between real-life consumer behavior versus perfectly rational consumers (whom we’ve yet to meet). We think we can dissect all the parts of ‘LIKING’ to scientifically show what contributes and what detracts from liking. To an extent, we can, but also to an extent, consumers just don’t work that way. They may just prefer a product (maybe the brand plays an outsized role, for example) that doesn’t result in all of the features laddering up to a nice neat score.
In sensory science, we use PENALTY ANALYSIS. This is a nifty tool where we can compare people who think there’s ‘too much’ of something to those that think there’s ‘too little’ of that same thing. We can look at how their OVERALL LIKING differed, which assumes a perfect science. And directionally, this could help inform how the different parts of the product made up the whole (sweet, salty, crunch, ratios of different components, etc). But we were often left a little confused as well. How could we get penalized to an equal degree for being BOTH too little and too much?
So back to the interplay between qual and quant learning. They both need to be used to inform a decision. Quant can only be as good as the questions we’re asking and often assumes extremely rational consumers (although different techniques like System 1 have worked hard to make less assumptions about rationality). Qual allows for the messiness of real life, but it can sometimes feel too messy to be useful. Keep in mind that consumer research needs to be a balance of BOTH art and science. Balance the math WITH the messiness. This is what makes our jobs as researchers both really hard and really interesting.
What to chat about how you can use consumer research to answer questions in your organization? Or how you can use both qual and quant to get a fuller picture? Schedule a call today!
