Recruiting people who are like your customers

The most damaging shortcut in research is the participant who’s easy to find. The colleague down the hall, the friend who runs a shop, the founder’s spouse with strong opinions about the website. They’re available in five minutes and cost nothing, which is why the study happens at all, and also why it tells you nothing about your customers. The person closest to hand is the person least like the stranger who’ll try to buy from you.

Convenience sampling costs more than wasted time. It produces results you’ll act on with confidence, and the confidence is the dangerous part. You watch your product manager use the checkout, they sail through it, and you conclude the checkout is fine. It’s fine for the person who helped build it and has learned not to see the broken bits. You’ve swapped a hard question, “can a stranger use this?”, for an easy one, “can the person who made it?”, and then trusted the answer.

The fix starts before recruitment, with criteria. Decide who the study is for, and say it precisely: a returning customer who shops on a phone, rather than “people who use the internet”. Then write a screener, a short set of questions that confirm someone fits before you book them, and use it ruthlessly. A screener is there to turn people away, and if it never rejects anyone, it’s a formality. The most useful question is one only your real customer would answer a particular way.

There’s a subtler bias in recruitment too. The people who sign up for paid research are a particular slice of your customers, the ones who want the voucher or who enjoy being asked their opinion. That’s fine as long as you know it and recruit against it, casting wide and screening hard instead of taking whoever replies first. Perfect representation is impossible, but you can get close enough that the findings transfer, and that’s what you’re aiming for.

Good recruitment is boring on purpose. Criteria written down, a screener that rejects people, enough participants that the first five stop sounding like a coincidence, and no colleagues, friends or founders with opinions. The version to avoid is the hallway test written up as customer insight, a few easy conversations with people who already agree with you. You’ll recognise it by the missing criteria, the missing screener, and a conclusion that happens to match what the team believed before anyone was recruited.

Put it to work

Write the three criteria a person must meet to be in your next study, then the three questions a screener would ask to check each one.

Your answer

Show the worked example

There is no right answer. Write yours first, then compare.

Sources

1. NN/g, Selection Criteria: How to Pick Your Participants
2. NN/g, Convenience Sampling: Avoid Bias When Recruiting Study Participants