When you need meaningful customer insights, would you rather talk to 20 customers, or survey 2,000?
Many people would rather survey the 2,000. There is something comforting about basing important decisions on big numbers. Heck, even I am comforted by large sample sizes. But sometimes that confidence is misplaced. For instance, think about what happens when you want to understand a very precise customer group.
Maybe you target IT managers who use a particular combination of systems management tools. Maybe you need to research beer drinkers who prefer local brews. How about high-income, single parents? If you need to collect customer insights from hard-to-find groups, surveys may simply be unrealistic.
Ideal Isn’t Always Real
In an ideal world, you would match your research objectives to the methodology. Need to discover emerging usage trends? Qualitative methods will be best. Need to measure interest in specific product features? Quant is the way to go. But standard rules don’t apply when the goal is to get deep, meaningful insights from a hard-to-find target audience.
And this why I have recently found myself advising not one, not two, but four different clients variations of this statement: “…for your needs, in-depth conversations with 20 or 30 clients will be more useful than doing a quantitative survey of 500 iffy ones.”
I hear the quant folks screaming as they read this. Can N=20 ever be truly “representative”? Of course not. But neither is surveying 500 people (or 2,000) that can’t be validated, give lots of neutral or “don’t know” responses, or otherwise give indications they may not be as qualified as needed.
An Issue of Trust
You and your colleagues won’t use market research data you don’t trust.
It’s that simple.
Hearing the words and answers of 20 clearly-qualified people engenders trust. You can tell they know what they are talking about; they can’t hide behind lots of “neutral” or “don’t know” check boxes. Their word choices are authentic.
In contrast, looking at a set of charts based on a survey of 500 people can easily raise red flags…too many “don’t know” responses, too many “neutral” responses. Other signs may emerge that make you think that some respondents should not have passed the screening criteria.
That being said, many will ask: What can you learn from just 20 conversations? Or 50? Or 10?
A lot. Especially if you ask the right questions, provide the right environment, and practice active listening. You may be surprised.