Inclusion surveys – five steps to success

As we previously discussed, it doesn’t make sense to focus on launching inclusion initiatives if you don’t have any diversity in your organization. On the other hand, diversity is meaningless if you don’t have an inclusive work environment. (You know – being invited to the meeting, but no one pays attention to what you say?)

Method-wise, measuring inclusion is similar to measuring engagement or work environment, but here are a few good-to-know things.

1. Follow data privacy regulations

You will most likely combine the inclusion survey with questions that concerns diversity. That means you will need to handle the process with care. (Make sure to contact your legal department for advice on your specific case.) Religious beliefs, sexual orientation, and ethnicity are data-points that are very sensitive to capture. Therefore you will most likely need to do a fully anonymous survey – as a contrast to engagement surveys that are confidential or pseudo-anonymized. You cannot use linked background information, and you will most likely need to send out a fully open URL.

2. Make sure there is evidence for what you ask about inclusion

As with all data-collection, make sure there is evidence in research for what you are planning to track. What “makes sense” when it comes to inclusion does not always guarantee you an effect. For example, the share of managers participating in an unconscious bias training might be a totally pointless metric. Even though inclusion is much about feelings, don’t only focus on surveys. Do your best to also track actual behaviors. For example, you can do the survey on how employees experience team meetings and in addition track who is allowed to represent the company on stage.

3. Slice n’ dice with care

The actionable insights are never in the overall data, but cross-tabulate with care! If you slice on gender and a specific ethnicity you might suddenly end up with very small numbers in your organizational data. This is dangerous from two aspects. First, differences between small groups might not be significant (existing in reality). Secondly, you might blow the sense of anonymity in the survey. Even if you report aggregated data, people might feel called out if negative insights are presented for a small segment that they belong with.

Don’t go into too much detail if you don’t need to, sometimes “norm” and “not the norm” is what makes sense to use from a statistical point of view to make sure that your analysis is correct. Collecting detailed data might give you a sense of control, but it also comes with lots of privacy issues.

4. Do you know what to do with the inclusion insights?

As with both inclusion, discrimination, and harassment, you need to have a plan in advance before measuring. What is your plan when you get bad anonymous results in return? You will not be able to ask follow-up questions afterward, so make sure that you ask for the context in the survey itself. Also respect that the ones that do not feel included, are not the ones that should be responsible for solving the problem.

5. Cross-connect to other surveys

You might not be able to ask about ethnicity in your onboarding or engagement survey, but you can add questions connecting to inclusion. That way, you can still follow inclusion on a broader basis. While not being able to collect all diversity data more than once a year, you need to go for the second-best option, and that one can still take you far!

The survey and the data can never be the end goal

Being able to track Diversity and Inclusion in a professional and structured way gives you a solid foundation for driving change. Note though, that even if you just read two long posts, they were still only on the topic of “how to measure it”. 90 % of the job is still there. As with all data-driven HR, once the analysis and storytelling is done, it’s up to leaders to make it a reality with support from facilitators, org-developer, and HR. The survey and the data can never be the end goal.

If you liked this blog post on inclusion surveys, you should check out Henrik’s post on Why your pulse survey won’t work.