Planning Poker: How To Make The Challenge Process As Transparent And Clear As Possible

    In the last post, we talked about how we work with the backlog , and today we will share details about the planning process, which in our case is not only useful, but also exciting, since we evaluate the tasks using Planning Poker.


    How and why do we plan

    Planning is a regular team discussion process for each task. We spend it every 2 days, and only members of the engineering team and those who set them the task participate in it.

    For planning, we collect all participants in a conference in person or remotely.

    The customer (task manager - product manager, marketer or even the general director) must be present at the planning stage to explain the essence of the task, tell how he understands it, convey to the team why the task is important and motivate the executors to complete it. The purpose of the engineering team is to find out the maximum number of details due to additional questions, draw out hidden requirements from the customer, offer their ideas for implementation and work out the best solution. Either explain why the task should be postponed for a while or not taken to work at all.

    Formulation of the problem

    The customer reads out his task (this is always done strictly by the customer in order to avoid a “broken phone” when transferring data and speeding up the decision-making process), explains what needs to be done and why it is important.

    The goal of each member of the team of performers , asking various questions to the customer and team members, is to find out what needs to be done and understand how to best solve the problem. Then the contractor explains what exactly he plans to do, and specifies whether this option will suit the customer and the team. All participants speak in turn, and after everyone has decided on the best solution, the stage of assessing the complexity of the task begins.

    Difficulty rating

    The complexity is assessed using digital cards - this is the so-called Planning Poker: all participants in the team of performers must evaluate the complexity of the task in days of labor, i.e. how many working days (based on the standard 8 hours) specifically for him (the participant) to complete the task. After all the performers have put the cards face down, everyone must turn their cards so that the numbers are visible to all participants in the process.

    Performers with boundary ratings (i.e., participants who gave the lowest and highest time ratings) should explain their choice. On the one hand, the pressure of the team will not allow participants to give an inadequately high estimate of the deadlines, on the other hand, the team gets the opportunity to discuss possible problems on the task. The one who gave the lowest score on time shares with the team exactly how he plans to complete the task so quickly. The participant who gave the highest rating should talk about the risks and difficulties he foresees in performing this task. After a second discussion, taking into account all the pitfalls, the team decides which assessment is more suitable for the task. This term is entered in the task description card in the task manager.
    The main goal of the complexity assessment is not to predict when the task will be ready, but to make sure that all participants equally understand the task.

    If one participant evaluates the task at ½ days, and the other at 3 days, they clearly decided to perform the task in different ways, and therefore should coordinate their actions and explain why they should do exactly the way they think. Sometimes the discrepancy in the assessment may be due to different experience in solving similar problems, in this case, the maximum assessment is taken, but if the time period differs by more than 1 day, then the task is performed in pair programming, when the one who evaluated in a smaller direction manages who appreciated the big.


    We must strive to ensure that the complexity of the task does not exceed 1 day.

    Planning is best done so that the team always has a supply of tasks for 1-2 days, but no more. So all participants will remember well all the details of the discussion of each specific task. If a task has hung in the queue for execution for more than 4 days, it needs to be tabled again so that the team remembers what and how to do on the task.

    If the task is performed 2 days longer than planned, it must be taken out again for planning and discussed with the team and the customer all the difficulties and problems encountered during the execution, so that the next time they are taken into account during the assessment.

    The Planning Poker format has well taken root in the work of our teams: developers, analysts, admins, and allows us to better understand the tasks and their solutions at the stage of setting.

    And how do you discuss tasks and evaluate the complexity and time of their completion? Share your experience in the comments!

    Also popular now: