How Planning Poker helps to Improve your Team's Estimations
Definitely worth a try, enjoy!
Reading Time: 7min
(www.robert-matthees.com)
(www.robert-matthees.com)
by Robert Matthees > UX > Planning Poker (2019-02-05)
Planning Poker (or Estimation Poker) is a method for estimating backlog items utilising the power of gamification. Every team member estimates the complexity of a task or backlog item in so-called story points by putting down a Planning Poker Card with an according number. Initially, everybody keeps her estimation hidden which means that the number on the card is pointing down towards the tabletop. This forces everyone to think on its own. Finally, when all cards lie on the table, every team member flips her card and reveals the estimation.
Now, the two team members who estimated the highest and the lowest complexity explain their decision. What is following is a quick discussion to negotiate an estimation for the backlog item that the whole team agrees on. Planning Poker works remotely too. Especially in agile environments like Scrum teams, Planning Poker quickly becomes the favourite approach. How does it help your business?
By introducing Planning Poker, you are enforcing a process of continuous learning. You may know a developer that always overestimates a piece of work or likes to go for a too complex solution (e.g. why using a standard sort function when you can write your own?). On the other hand, there are developers that tend to underestimate tasks and won't get their work done on time. Planning Poker helps here as estimation isn't a single person's decision anymore. The team members with the smallest and highest estimate will explain their decision and in the end, you will have an estimate that the whole team agrees on. Like this, Planning Poker is enforcing knowledge share too.
As two team members with the highest and smallest estimation will explain their decision, the team will see what makes the solution look that complex or simple. Maybe one knows a class, library or plugin that would reduce the work to a minimum? Don't miss these ideas anymore. Planning Poker will make best practice visible.
Further, you will see if there are any open questions about a backlog item during the discussion. Like this, Planning Poker will help you to clarify questions with regard to the scope and value of a backlog item upfront that they won't distract or negatively influence the actual development process later anymore. Pro tip: Have your Product Owner available for open questions.
All this usually results in less work for the team leader as it empowers the mindset and all the behaviour patterns that are key for self-organising teams: an upfront approach, individual commitment, transparency, quick adaption and open inspection.
Every team member gets a set of Planning Poker Cards. I like to work with a set of simplified Fibonacci numbers (see the explanation of different card sets below).
Now, the Development Team Leader or Moderator will introduce the first backlog item that needs to be estimated.
Then, every team member will pick an estimation from her set of Planning Poker Cards. The card gets put on the table with the number facing down to the tabletop that no other team member can see her estimation. When every team member has put down a card, everyone reveals her estimation by flipping her card. Like this, Planning Poker forces everyone to think on its own and not just to put down the same estimation as another team member.
Once all Planning Poker Cards are revealed, a team member with the smallest estimation and a team member with the highest estimation will explain their decision. Afterwards, the team will discuss the different points of view quickly until an estimation is reached that the team agrees on. It's helpful, to work with time slots here (e.g. 30 seconds speaking time). The moderator or team lead will take notes that are valuable during the development process later.
Once agreed, the next backlog item follows. You can do it for all or just for the larger items of your backlog. In order to increase transparency and to establish a process, of course, it's better to do it for all items and user stories.
There are different types of Planning Poker Card Sets and even Planning Poker Card Apps out there. Furthermore, there are digital solutions and several agile project management tools like JIRA support Planning Poker too (see comparison list ).
Very often, Planning Poker Cards contain a sequence of Fibonacci numbers. The Fibonacci sequence is a sequence of numbers where each number is the sum of the two preceding ones (e.g. 2 + 3 = 5; 3 + 5 = 8; 5 + 8 = 13). This is causing the gaps between the numbers to become larger the higher the number becomes. Such a sequence is a perfect fit for Estimation Poker as estimations become vaguer, the larger the complexity gets.
Often, a set of Planning Poker cards will contain a question mark, an infinity symbol and a coffee cup too. As the numeric cards stay for story points (complexity of a task or backlog item), the question mark indicates a value that is not known. Here, the backlog item or user story needs to be discussed more.
Whereas the zero indicates a task that is already done or a backlog item that nearly contains no work at all, the infinity symbol indicates a never-ending story or task or a much too big backlog item that you need to break down into smaller user stories. If the Planning Poker Card Set doesn't contain an infinity symbol, you can use the question mark as a synonym. I rather prefer these sets as it reduces complexity.
You will throw the coffee cup when you need a break. Yes, this is the most important card of your set!
There are other, non-numeric Planning Card Sets out there too. Personally, I prefer to work with a set of simplified Fibonacci numbers as explained below.
(Screenshot, Scrum Poker App by artArmin from Amsterdam)
This is my favourite set of Planning Poker Cards. It contains a simplified Fibonacci sequence from 1-13, followed by the numbers 20, 40 and 100. From my experience, this set works best as it reduces the complexity of Planning Poker. Just compare it with the Planning Poker Set below and see the difference to a set that contains a full Fibonacci sequence.
(Screenshot, Scrum Poker App by artArmin from Amsterdam)
This Planning Poker Card Set contains a full Fibonacci sequence. Logically, it makes somehow more sense. But in my eyes, it makes the game of playing Planning Poker much too complex.
(Screenshot, Scrum Poker App by artArmin from Amsterdam)
Here you see a set of cards using the traditional T-Shirt Size Estimation. You will better go for story points as they give a much better indicator for complexity and development work as traditional T-Shirt Size Estimation or guessing time and man days.
(Screenshot, Scrum Poker App by artArmin from Amsterdam)
This set of Planning Cards consists of different colours and helps to estimate potential risk.
How to start? Take the smallest or a suitable small item of your previous sprint and it gets 0.5, another item gets the 1. From this, you start your estimation. By practising Planning Poker, you and your team will get a feeling of proper estimation and of how much story points you will get done during a sprint.
You can organise Planning Poker easily via Slack or Skype too. Every team member sends in her estimates in a private chat dialogue to the moderator. Once all estimates are in, they get revealed and the discussion happens in the group chat. Of course, it's not as nice as estimating with real Planning Poker Cards, it's even worse than a Planning Poker App. And, unless you have video chat, you won't see your peer's Planning Poker Face, but it's a good solution to involve remotely working team members as they can join your Planning Poker session from home office. Maybe, you will even find a set of Planning Poker Emojis to enrich their experience.
Planning Poker usually results in a much more self-organising team, higher commitment and better knowledge share. Definitely worth a try, enjoy!
"Living in a world of products, making it more fun!" (Robert Matthees)