Agile Scrum Masters can create better sprints faster using
Situation: The next Agile sprint planning meeting is coming up and there are a dozen stories at the top of the backlog. How do you select the stories that have the most business impact and maximize the time spent developing code in the Sprint?
Challenge: Agile teams are multi-disciplinary groups of people that need to work faster and smarter. The development team wants to avoid sitting in meetings and devote all of their time to developing code. The business leaders want to be sure they are building the most impactful features. Coming to an answer for the Sprint requires consensus from the business and input from the developers, and it needs to be done without endless meetings and deliberation.
Traditional Solution: A typical approach for planning is to meet with the business owners to rank the stories from most important to least important. The prioritized stories are given to the development team to point (i.e., estimate the time it takes to complete each story). Once the stories are pointed, the scrum master creates the Sprint by starting with the most important story and he continues down the list deducting the point estimates from the budget until the budget is exhausted. This approach is flawed because it ignores the relative value of stories. For example, Story A may be twice as important as Story B but cost three times more to develop. It’s also possible to have ‘big’ stories crowd out lots of ‘small’ stories that cumulatively could have more business impact than the larger story.
Let us show you a better way using GridRank!
Create an Agile Planning Game
Select Budget (Pick Many) from the drop-down, this tells GridRank the game type, which we refer to as a Budget game or Pick Many, because the GridRank’s algorithm will try to pick as many stories as allowed by the budget constraint. Next enter the name of the game. Next, because this is a game about Agile, we are going to change the Options Label field to “Story.” Finally, enter a budget number into the Budget field. Think of the budget as the total number of man-hours that the development team has to spend for the sprint, which in this case is 160 hours (a two-person team with each team member working for two weeks at 40 hours per week).
Enter your Evaluation Criteria
Evaluation criteria are the factors that each player will use for establishing the importance of each project. In an Agile Planning game, it is common to have only one criterion — Customer Impact. Customer Impact and its axes labels are entered in the screen below.
Enter your Options
Enter each of the stories that will be ranked in the game. Budget Games require that you enter the cost for each story; in this case, these are the point estimates from the development team. You can also select a color and upload an image.
Click here to play this actual game now. After you play, come back here and we’ll discuss some of the behind the scenes calculations that GridRank uses to develop optimal results that are better than the traditional method for sprint planning.
Behind the Scenes
GridRank uses linear programming software to optimally determine the set of options that best meet the evaluation criteria, staying within the 160 hours constraint. In other words, once the players have submitted their plays, GridRank determines the set of user stories with the highest possible value score subject to the 160 hours budget. In the table below, we show the results table from the game where GridRank determined the “In” stories. GridRank selected six stories — these are denoted “In” in the far right-hand column. Further, if you add up the $Cost and %Desire scores for these six, you’ll see that GridRank stayed within the 160 hours budget with 4 hours to spare, and it maximized the total %Desire score to 68.1 out of 100.
GridRank is Better than the Traditional Approach
The traditional method of Agile Planning involves ranking the user stories from most important to least important, and then selecting the stories until the budget is exhausted. The table below demonstrates this method. Using the traditional approach, the scrum master would select the 1st, 2nd, and 3rd stories with highest %Desire scores (i.e., highest priority). Selecting these projects exhausts 140 hours of the budget. The scrum master would go down the list, looking for a project for 20 hours and find he could add a Hover Description, thereby closing the budget.
Which Solution is Better?
The total %Desire score for these four projects is only 54.4 which is 20% less than the GridRank Solution. Further, the Traditional approach exhausted 160 whereas the GridRank solution spent only 156 hours.