The Clover 5: Measuring employee happiness

We often describe Clover as a behavior change engine: our job is to use data, science, and intervention design to help our members be as happy and healthy as possible. But to do that, we need to have a happy, healthy workforce. So we use the same process with which we change behaviors externally to work on our internal culture and environment. Because to create a workplace that provides the opportunity to do meaningful work, we need the same rigorous adherence to science.

To do good science, you need an outcome measure. So at Clover, for about the last year we’ve been collecting something we call the Clover 5, a short measure based on the science of what makes people happy and productive at work.

While the wording has pivoted around a little as we’ve grown into the survey, the current version is as follows:

  • Do you like who you work with?
  • Do you trust who you work for?
  • Are you good at what you do?
  • Is what you do meaningful in Clover?
  • Is what Clover does meaningful in the world?

Our collection methodology is simple. Each week, we survey a quarter of the company, which avoids survey fatigue while making sure we have a continuous data stream and get to hear from everyone each month. The monthly scores are aggregated and reported on a dashboard that is open to everyone at Clover, along with participation rates and an overall average score.

For each question, employees can also choose to leave a comment. While these are not published, they are reviewed by our HR team and then boiled up into an insights report that is discussed monthly at the Management Team meeting. Each insight is paired with an appropriate verbatim that contextualizes it and then the Management Team discusses action items that follow.

We’ve also recently added a free-text question asking “Do you have any suggestions on how we can work together to improve any of the Clover 5?” If appropriate, those are posted alongside the Clover 5 data, along with a directly responsible individual that is taking them on and a continual tracking of progress.

But what about the numeric data? For us, the Clover 5 represents an outcome measure, not a process measure. So it can’t tell us what to do, only whether the pilots we run are working and what should be tested and scaled.

So, for example, consider the perennial “Is leadership training worth it?” The traditional challenge has been defining what “worth it” means but the Clover 5 can help us answer that by providing a benchmark against which we can measure ourselves.

When Clover introduced our first round of leadership training sessions, we looked at this exact question. First, we analyzed which managers had attended the trainings, then compared their team scores against the Clover average to compute a difference score. We then looked at this difference score before and after training, effectively comparing their untrained selves to their trained selves.

The result? Leadership training had a measurable positive effect on the Clover 5 score of the teams whose managers had been trained. But there is a wrinkle. Our leadership training was voluntary and the managers who opted-in to taking the training were already outperforming the average Clover manager before they took the trainings.

So while we know that our leadership training makes good managers better, what we don’t know is the effect of training on managers whose teams report lower Clover 5 scores. But because we have a score, we can easily find out.

We’ve used the data to evaluate many other interventions, from the effect of an Appreciation Week that celebrated the contributions of a key part of our organization (average score up!), to the effect of new management team members on their orgs (depends on the management hire!), to the effect of communication around strategic priorities (depends on the team!). We’ve also looked at predicting behaviors like attrition, promotion, and performance to see where the scores on individual questions can help us understand where to focus.

And perhaps most excitingly, we added the Clover 5 to our recent Inclusion Survey. This will let us compare the overall working experience of different demographics within Clover to ensure that we are striving for truly equitable environments. Because inclusion is more than just about fighting the hiring, pay, and promotion gaps, it is also about the day-to-day life of our employees.

This is the first time we’re talking about the Clover 5 externally in a formal way, so hopefully you’ll excuse the focus on our process rather than our results; because we’d love to see other companies adopt similar techniques, we wanted to get the method out first. In the future, we hope to publish on a more regular basis about the experiments we’re running using the Clover 5 and encourage other companies to start doing the same with their own internal metrics.

Teddy Roosevelt said “Far and away the best prize that life has to offer is the chance to work hard at work worth doing” and we believe that by everyone sharing how they fulfill that vision, we all can create better workplaces.