Monday, December 27, 2010

Scrum Inc Sprint 2 Retrospective: The Happiness Metric

Happiness Metric Process Improvement Priorities - ScrumInc 15 Dec 2010

Scrum, Inc. was a small company hosted by OpenView Venture Partners in Boston from 2006-2010. As a two person Scrum using Pivotal Tracker as a Scrum tool, I was a remote Product Owner traveling 2-3 weeks a month in Europe. Operating remotely was an impediment and Scott Maxwell, the founder of the venture group has been encouraging me to expand the business for several years.

Effective 1 December, we are now six people. We had a workshop on 6 December to do initial planning and backlog creation and we run weekly sprints. When I returned from Europe mid-December we had our first sprint planning with the whole team colocated.

We decided to run the weekly sprints using a format similar to that used by Scott Downey at MySpace and Napster. He runs the Sprint Review, Retrospective, and Sprint Planning as all one meeting that takes half a day. It took us a full day as we introduced the Happiness Metric as a way to run the company and had a lot of Product Backlog grooming to do.

We decided to use the introduction of the Happiness Metric as the Retrospective for the previous sprint. The Happiness Metric is a leading indicator for revenue and was discussed in a previous posting. The retrospective can be done in many ways and consists of several key components:

1. Establish a clear picture of what happened during the last sprint - the good, the bad, and the ugly.

This we did by having everyone answer the first two questions of the Happiness Metric - on a scale of 1-5 (1) how do you feel about the company and (2) how do you feel about your work at the company?

Everyone put a number on two sticky notes and posted them on the wall. The average for the team was little over a 3. Not great, but not bad for a team that was just starting up and several of them were learning Scrum for the first time.

2. Create a list of process improvements that emphasize the good, deprecate the bad, and pretty up the ugly.

We did this by asking the second two questions of the Happiness Metric - (3) What would make you feel better about the company and (4) what would make you feel better about the work you do?

Each person created one item per sticky note and put several notes on the wall.

3. Identify the top priority process improvement.

It is critical to get extreme focus on the process improvement that will have the most impact for the least cost. Fixing one thing at a time gets everyone behind making one change. The team can then feel the impact of the change and discover whether the change worked. Following that, the priorities of process improvement may radically change. The next priority process improvement can then be selected in a way that maximizes progress and minimizes wasted time and effort.

We decided to use Priority Poker to rank order the improvements - the things that would make people happier. We pulled out our Planning Poker cards and to my surprise, improving the way the team worked together scored 64 points, way above anything else! Everyone was astonished as we would never have seen this by just having a conversation about what was important. When we drilled down into how we would measure this during the next sprint we discovered several things were important:

a. People were working only on their own stories and need to start working together on stories. We will measure how much people team up on stories.
b. People who should be in the loop on discussions need to be included. We will measure how many times this happens or does not happen.
c. Avoid duplication of effort by including the right people in the discussion and having daily meetings at the same time and place. We will measure the number of times miscommunication causes rework.
d. The sprint must have points set aside for unanticipated events. Any new items coming into the sprint in this category must go through the Product Owner. If the Product Owner finds it necessary to insert more new work than points allocated, the sprint will be aborted and replanned.
e. Pivotal Tracker must be updated at all times so that those not in the office can see the global status of everything in the company.

4. Implement the top priority process improvement before the end of the next sprint.

A lean expert was in our Scrum training in Paris the previous week and his critique was that we do not emphasis Kaizen enough. Work to be done = Product Backlog + Kaizen. In Scrum, Kaizen is impediment removal. We decided the top impediment in the retrospective would always be a story in the Product Backlog for the next sprint. The number of priority points for the top impediment was far larger than the business value of any product backlog item.

Of interest, is that the second priority was getting new space for the larger team. A backlog item was put into the sprint to investigate this. Within two days we have several candidates and our new lawyer from the venture community advised us to locate at the Cambridge Innovation Center as this is where all the action is in the Boston new venture community. Even though this effort was lower priority we negotiated new space before the end of the sprint!

The rest of day was even more surprising. My feeling, based on work with high performing teams, is that this team will go hyperproductive and if they do, will be the first office Scrum that has the metrics to prove it. We broke out champagne at the end of the day and celebrated the launch of our first properly planned sprint. My happiness metric was already rising and I was looking forward to our next meeting!


Henrik Kniberg said...

Glad to see you're putting the happiness metric into practice!

Henrik Kniberg said...

Glad to see you're putting the happiness metric into practice!