Estimating has always been closer to a dark art than a science. But the more you can apply practical, repeatable process the better the estimates will become. A good estimate is based on solid data; any gaps will throw out your total. So let’s examine the common problems seen when estimating project effort and timescales. It’s so much easier to estimate how long something might take once you have done it. But often you will face the situation of creating an estimate for something you have not done before.
Collecting new information and re-using historical data
So when starting an estimate, the more information you gather the better. If you are working on something totally new to you then find someone that has experience of this type of project before. If you do find yourself in the more extreme situation of doing something that has not been done before then look for something the is a reasonable match to establish a basis for estimating. Remember to save the estimates and actuals from this project and re-use data from previous projects. It is amazing how few project managers and organisations have historical data on previous projects. As a result, many organisations have issue with estimating and in many companies fewer than 50% of projects are on time.
Detailed and Coarse estimating
Remember the old project management adage about “how do you eat the Elephant?” the answer is “one meal at a time”. Therefore, in estimating breakdown you project elements and estimate at the lowest level for work that will be done it the short term. This is good practice regardless of your project method. But how far should you go in breaking things down, that will depend on the nature of the project. While you want a fine level of detail, you do not want to make your estimating and planning so complex it becomes unusable.
Once you have a level of detail the provides short term control, prepare detailed estimates against those tasks that will take place over the next few weeks. Then use the summary information from these tasks to produce a coarse estimate for later months and longer term into the project. You should do this because why should you spend time producing detailed estimates for later dates, when that is likely to be modified by what you learn from those early tasks.
Tuning estimates to adjust plan and ongoing forecast
There is a peak of estimating at the start of a project, consequently new project managers assume that is the end of estimates. In some cases, project staff pour a great deal of work into the estimate to build a plan to compile or confirm a budget, never updating the plan beyond that point. The plan and its underlying estimates are only ever a model of the real world. The plan is unlikely to be an exact model. But the closer you can get your estimate and plan to reality the better you chance of success. As your plan progresses track the plan to establish your current status. Use the variation between the actuals and the estimate to tune the future plan. Now record the estimate and results for all your projects. Over time you will see improvements for future work.