How Many People Should Be Responding To My Research Surveys?

When leading innovation, youÕll want to have a system for doing research. Research doesnÕt have to be in the form of surveying customers, but asking a potential customer if theyÕd purchase a new offering is typically done at some point in the innovation process.

innovation and consumer research

The question we get asked quite frequently is how many customers do I need to survey. My answer is that it depends.I know, not the answer youÕre looking for, but itÕs the truth. The problem is that you need to plan and budget for future surveys about ideas that havenÕt been created yet. It depends, isnÕt helpful, so here are two things to consider:

  1. How confident do you want to be? The more confidence you want to have in the surveyÕs results, the more people you should survey.
  2. How much variance have you seen in past surveys? This canÕt always be applied, because when leading innovation you will have to test new ideas with new customers. But the take away is that the more variance the questions have, the more respondents you want.

With estimates for confidence and variance you can use statistics to calculate the exact number of responses you should be collecting.

Bigger picture. I recommend having a system for testing ideas throughout the innovation process. With new ideas, test lots of them with small sample sizes. You donÕt need a lot of confidence here, just enough to justify working on an idea. Systematize this testing, so there is always time and a budget to test new ideas. As ideas get more defined and are ready for development you can retest them with larger sample sizes. This takes more effort, but youÕre doing this with less ideas and gaining the confidence you need to give the go ahead for development.

Your job as your organizationÕs leader is to create a system that enables your employees to collect the responses they need. Organizations that have a system for research surveys will also have the data needed to calculate sample size.