One of the most difficult and consequential steps in a project is effort planning. For this we want to give you some proven rules here:
- To help you make a reliable effort estimate, you should know relatively well what you need to deliver; otherwise you can only guess. Not knowing exactly what to deliver may be because you have no experience of what you are getting into. It can also mean that your customer does not really know what he wants.
- If you have never created or delivered something similar, you risk being far off the mark with your effort estimate. There is no substitute for experience, not even the method described below.
- If you think that this time is much faster than last time, you should have a sober reason. New tools and methods are not included.
- In some areas of technology, the performance of the individual team member plays an essential role. Especially in the field of knowledge work, productivity differences of 500% are not uncommon. Except for very large teams, where these differences are clear, you should therefore know your team members and include them in the effort estimation.
- Do not downplay the effort of your team members. As a rule, your employees underestimate the friction that occurs and overestimate their productivity.
Planning poker is an interesting method to come to toughening assessments after paying attention to the above points. Prerequisite for a planning poker is a list with the project objectives. These must be described so accurately that the estimators can understand what the project is about.
For planning poker, the estimators meet under the direction of a moderator. Each estimator receives an identical stack of cards that you can buy or create yourself. The cards contain for example the numbers 0, 1/2, 1, 2, 3, 5, 8, 13, 20, 40, 100 and a ? for the statement that an appraiser does not feel able to give an estimate. The numbers can mean the effort in days, or in another, previously fixed constant relation to the real effort.
At the beginning, a product manager presents the project results to be delivered, as it has been defined in a project management software like Allegra. The team can ask if something seems to be unclear. Then each team member puts the card face down on the table with his guess. When everyone has discarded his card, the cards are turned around at the same time.
The estimators with the highest and lowest estimates are given the opportunity to explain and justify their estimates. The group can discuss the arguments. The discussion time can be limited by the moderator or project manager. The estimation process followed by discussion is repeated until the group has reached consensus.
The estimator responsible for delivering a partial result always has more weight in this process than the other participants. The numbering of the maps takes into account that the larger the estimate, the greater the estimation inaccuracy. The estimator can thus not pretend any supposed accuracy for larger values, but has to decide on a pessimistic or optimistic value.