Monday, November 29, 2010

Done-Related Story Estimation

In today's Backlog Grooming I tried a new method of estimating user stories. The estimations were based on and related to known done-stories of former sprints. The method is very simple and contains only five steps. 

Step 1: Get user stories from Product Owner

Your Product Owner should be prepared for Backlog Grooming and bring his top-priority stories for discussion and estimation.

Step 2: Select reference stories

Pick a few random stories from your already done stories. Make sure their story points differ enough so a typical mixture of story points is in the set. In today's experiment I chose stories with 3, 5, 8, and 20 points.
Then write a new reference story card for each of them in the following way: write the story on the front of the card and write its story points on the back of the card. Make sure to hide the story points from the team's eyes for now.

Step 3: Sort user stories

Let the team find the right order of old reference stories and new undone stories. You will notice much more comparative conversation on the stories than with playing Planning Poker.

Step 4: Show story points

Now the fun part begins. Turn the reference stories and surprise the team by showing the reference story points.

Step 5: Adjust stories and story points

Let the team discuss if any adjustments are necessary. There will be more conversations on single cards to gain more understanding. Some stories may change due to these conversations. Take care that stories with high points get split properly. In today's example the story at the bottom with 20 points could be discussed and split into two stories. The result was much deeper understanding of the real value for both the team and the Product Owner. 

I'm looking forward to the Sprint Planning meeting next Monday.


  1. Hi Marc,

    first let me thank you for sharing your idea. Great timing since I was looking for something to spice up the estimation meetings for the teams.

    The reference cards worked like the "Gather Data"(What happened in the Sprint?) part of a retrospective. The groups started to discuss the reference stories, what got things started and continued with the new stories. There was a lot of energy in the air. Let me add, I split the 2 teams in 4x3 groups.

    Turning the reference stories every single group had the right order of story points, what led to a surprisingly and nice ending of the meeting.


  2. Hi Marc
    I got very inspired to try this estimation method instead of the size estimate "played" with poker planning.

    Please continue up this great job of sharing :-)

    Jakob Schmidt Sørensen

    You might what to read me blog http:\\ (I just begun blogging)

  3. Hi there, awesome site. I thought the topics you posted on were very interesting. I tried to add your RSS to my feed reader and it a few. take a look at it, hopefully I can add you and follow.

    Function Point Estimation Training