Measuring Happiness

There are lots of posts around metrics in agile. In this post I want to talk about a different kind of metric: Team happiness.

The post was inspired by a great talk at SPIN last night by Sharna Sammy. She talked about how her design team in the publishing world used a Niko Niko calendar to track their team happiness. They created a cool template for each person to track their own emotions daily. First they started tracking them individually, then later they made it visible on their Scrum board to the rest of the team. Sharna mentioned how even though it is not a number metric, there were definitely interesting patterns that could be seen. The whole team was on a low when they were working on stuff they didn’t enjoy.

You might think that’s obvious and wonder why you should care. We were talking to Alex Kjerulf (Chief Happiness Officer at Woohoo Inc) when he was in South Africa a few months ago. He talks about happiness at work. One of his findings it that people are happier at work if they can acknowledge and talk about their emotional state as well as just progress on projects. Apparently being able to mention that a certain project is getting you down is a key element in improving happiness at work.

So what can you do? I’ve used 4 different ways to measure this with various agile teams.

Daily score out of 6.
In my first agile team doing XP, at the daily standup we would each give a score out of 6 each day for what our happiness/energy level that day. Why 6, well it forces you to choose something other than the middle. I find if you use 5, people often pick 3. I tracked the average of this number each day, and could definitely see peaks and troughs and understand what caused those. It also helped the team. If someone was having a bad day, they would say 1, and people would know not to bug them too much that day. If someone said 6, it inevitably made others feel good too.

Niko Niko calendar
As mentioned by Sharna, I’ve also used a Niko Niko chart on the team board for people to draw a face about how they are feeling today. This works in the same way as the daily score, except it’s a little harder to track a number metric from it (like an average), but it has the benefit that people can be a bit more expressive. When I’ve used this I haven’t tracked the information over the long term, just made it visible in the sprint for each day.

Net Promoter’s Score (NPS)
NPS is usually used to measure customer satisfaction. It’s a great metric because it’s easy to measure. You ask customers a single question. “How likely are you to recommend our product to a friend?” They score out of ten. Scores of 6 and below are counted as detractors, scores of 7 and 8 are neutral, scores of 9 and 10 are promoters. You subtract the detractors from the promoters and divide by the total number of people who answered to get a percentage of promoters. This is the percentage of people who actively promote your product/brand.

For example:
If you survey 10 people.
2 say 5 out of 10
2 say 6 out of 10
1 says 8 out of 10
3 say 9 out of 10
2 say 10 out of 10

Then you have 4 detractors, 1 neutral and 5 promoters.
NPS = (5 – 4)/10 = 10%

I turned this around and asked team members how likely they would be to recommend working at our company to a friend. I tracked this once a quarter at the start of our release retrospective using anonymous ballots. Download my xls to print out the ballots and calculate your score. I used this as a dev manager to measure how I was doing at creating an great environment to work in. The score is not as important as the trend. You want to track if this measure is going up or down over time. I didn’t share the data with the teams, I used it as a metric with my boss to see how I was doing as a manager.

Gallop 12 – Employee Engagement Survey
I’ve also used the Gallop 12 survey. This is a list of 10 questions for people to answer for themselves. I’ve used this in a retrospective for people to score themselves and then discuss how we might improve engagement. I never tracked what this was across the team, but some people willingly shared their scores. What was more useful was the discussions once people realised some factors that might be missing for them in the work place.

A word of caution
One thing about happiness metrics. These are lagging metrics. They won’t tell you immediately if there is a problem. People’s mood only gets worse/better on average a while after changes that impact them. However they do tend to be an earlier warning sign than waiting for staff churn. People are generally unhappy for a while before they leave. I found a direct correlation between NPS and staff leaving.

Explain to people why you are measuring it. They need to understand the consequences, and there needs to be no consequence that might lead them to game the metric. For example praise if everyone is in a good mood, or ‘motivational’ sessions if everyone is in a bad mood. Nothing sucks worse that a ‘motivational’ talking to if you are having a bad day.

Hope you enjoy, please share in the comments if you use any of these ideas, or have others. Thanks to @SharnaSammy for a great talk and to @NateKettles for asking a question on twitter that inspired this post 🙂