A question that I’m often asked is: what’s a good participation rate for a developer experience survey?
For medium to large tech companies, the average seems to be around 30%. I’ve seen numbers closer to 60% for companies that use methods such as sampling a subset of their developers each quarter.
At DX, we’re proud of helping our customers achieve average participation rates of above 90%. We’ve sustained this over several years with companies ranging from hundreds of developers to several thousand.
Here are some of the lessons we’ve learned about the most common causes of suboptimal response rates:
Surveying too frequently. The most important rule of surveys: don’t run them more often than they can be acted upon. Many leaders are driven by a desire for continuous data points and overestimate their teams’ capacity to act on survey results. A common mistake I’ve seen is where organizations try to run monthly surveys even though teams are only able to commit to new initiatives each quarter. It’s frustrating for developers to be asked about the same issues repeatedly when action can’t be taken.
Unclear purpose. If someone came up to you on the street and asked you to fill out a questionnaire, you’d need a compelling reason to do it. Many developer surveys I’ve seen are rolled out haphazardly without clear communication. Who exactly is the survey for, and how will the data be used? Be clear about your purpose, then make sure you follow through.
Poor survey design. Nothing says “this is a waste of time,” like asking developers questions that are confusing or difficult to answer. This problem is easily avoidable through rigorous survey design methods, but many organizations don’t invest adequately in this.
Lack of personalization. One common survey limitation is the lack of personalization for questions that are asked. For example: developers being asked about tools they don’t use, or QA engineers being asked about CI/CD processes that aren’t relevant to their work. For a deep dive on survey personalization, listen to my podcast episode with LinkedIn’s developer insights team.
Insufficient follow-up. Surveys need to result in positive change or effort. Otherwise, what’s the point? There are many different approaches for following up on surveys, but it’s ideal when leaders from multiple levels of the organization take part in the effort. Executives should be focused on addressing cross-cutting issues, and front-line managers should address issues that are specific to their local teams.