This post introduces a technique, which we have labeled “3D Planning Poker”. It is aimed at helping agile teams with their user story estimations. It was invented by Jannik Streek, who is coauthoring this post together with me. At first we have a look at estimations, mentioning our perceived short comings of Planning Poker and then we introduce our suggested technique.
Estimations and (normal) Planning Poker
Estimation of user stories is an important task when developing software. It is especially well known and practiced in agile software development processes. Teams rely heavily on estimations during their iteration or sprint planning meeting in order to commit to a reasonable amount of work during the next sprint/iteration. Obviously false estimations can be a huge problem. Therefore good estimations are desirable.
Due to this reason a couple of methods were invented, most prominently Planning Poker. Planning Poker provides an easy estimation method which gets everyone involved. Moreover you get the unbiased estimation of every team member since no one knows what the others will estimate when they show their own estimations. However in our minds it suffers from a couple of problems:
- Quietness: if you have people in your team who are not that extrovert, you will probably have the problem that they will not be that active in your discussions.
- Late feedback cycles: You will only get adjusted estimations after a re-estimations round which also takes time. Often it would be beneficial to get more direct feedback.
- Tiredness: Usually estimation is kind of a boring process: Sitting around and estimating can be time-consuming and dull. Thereby tiring the whole team. Moving around the room can be done during breaks, but maybe it would be even better to include that in the estimation process?
- Slow to get the big picture: At first you have to look at every card of your team members in order to recognize what the team estimated overall. Cards have to be shown around because not everybody can see them (the card of your neighbor is hard to see, for instance). It takes some time before you know what everyone estimated.
Introducing 3D Planning Poker
We want to introduce 3D Planning Poker in order to address these short comings. It’s a technique / process which was actually developed during a course in Software Engineering at the Hasso Plattner Institute. The basic idea is to have an estimation method with “live” feedback and movement to don’t get tired during the estimation. At the same time we experienced that even rather introverted people get more motivated and active.
The technique works like this: In your room you put your usual estimation numbers (for instance: 1, 2, 3, 4, 5) up as big numbers on the walls. Make sure that the numbers are in order. At first a user story is discussed, as usual. But then when you are actually estimating everyone should just move to the number on the wall matching his/her personal estimation. Everyone should move at once so people don’t get biased when they see where everyone else already stands or in which direction co-workers are heading. When everybody arrives at “their” estimation number you can easily see where most people stand and how far apart people stand. That gives you a nice visualization of the estimates and thereby an immediate impression of the current “estimation situation”. The people with the biggest difference in their estimation also have the biggest (room) distance between them. Just like with normal Planning Poker the people with the most diverging estimations should explain to each other why they estimated the way they did.
A really cool thing about this technique is the live feedback. People are allowed to move to new estimation numbers during a discussion, which gives you immediate feedback. Plus: there is no need for a re-estimation as the team estimation changes dynamically. In our meetings it was the norm that during the discussion people got convinced by an argument or simply noticed that they misunderstood the story, so they moved into one direction or the other. Sometimes even one of the people discussing would get totally convinced and moved to the spot of the other disputant.
The only concern regarding this method is that people might be influenced by others, because they see them walking to an estimation. So Charlie could always try to follow Steve to his estimation. But this bias is observable, as the team can notice when someone always follows other people or always stands next to specific persons. You lose the total unbiasedness of Planning Poker but you get some more movement, live feedback and possibly more engagement.
In order to achieve the best results you should have a team in which everybody trusts each other and actually has the courage to choose a number by him- or herself. That’s why we believe that this technique might be unsuitable for freshly formed teams. However we encourage you to try it with more mature teams especially if you notice that planning meetings are becoming a “boring routine”, which they should not be.
What do you think about 3D Planning Poker?
We shared this idea with some people at the XP 2012 conference and got very positive feedback, which actually motivated this blog post.
What are you thinking? Please share your opinions, ideas and/or concerns with us! Do you have other kinds of techniques or processes in your team for estimation that you would like to share? Did you try 3D Planning Poker and would like to share your experiences? Please comment!
Jannik & Tobias