Our Project Estimating is Waaay Off. Top Tips for Success.
In The State of Project Management Report 2017 one of the identified largest challenges identified by organisations was planning. A core component of planning is good estimating. 63% of organisations from the same report highlight that projects finish late half the time or even more frequently.
Asking colleagues to estimate can sometimes be met with “I don’t know until I do it”, “I don’t have a crystal ball” and my favourite, “how long is a piece of string”, as long as you want, if you are in control, have the scissors and make a decision!
So here are my top tips to improve your project estimating. Of course these apply not just to time but also to budget and resource too:
1) Best practice guidance
The Chartered body for our profession, the Association for Project Management (APM) cite 3 techniques for estimating:
- Comparative estimating uses historical data from similar projects to extrapolate estimates for new work. Past estimates are scaled according to scope and complexity to produce new estimates
- Parametric estimating uses defined parameters by which work can be measured. For example £30m per mile of motorway
- Bottom up, or analytical, estimating uses a detailed specification to estimate time and cost for each component of the work
By its very nature, estimating is speculative. It involves looking into the future and deciding what will happen. The more data available, the better quality the estimate. In the early phases of a project less information is available. There are more unknowns and questions. This results in the Estimating Funnel, as more information becomes available, then the quality of the estimate improves.
2) People tend to underestimate
The majority of people underestimate. There are of course exceptions, but let’s stick with the majority. Why do they do this? Do they want to sabotage your project? No! In fact people are trying to demonstrate how good they are at their job. “How long does that report take?…oooh, I can do that really quickly, I’m good at my job”. This typical thought process leads to an unrealistic estimate, something way too positive. You might be impressed with their estimate, but not reality when they don’t deliver. So take a step back, and a more structured approach.
3) Ask for how many hours over what elapsed time
Rather than asking for how long do you think this will take, let’s break this down into two very specific questions. Once we are all on the same page in terms of the scope (itself not to be ignored) then your first question should be about work hours “How many work hours does this task take?”. Once you agree on the estimated work hours, then ask the second question “Over what elapsed time can you undertake these hours on this task?” For example if the work hours estimate is 8 hours are they free tomorrow to work on this task tirelessly and complete it over the elapsed time of 1 day, or are they only going to be able to allocate one hour a day except Fridays so it actually takes two weeks! Where you want to get to as a Project Manager is agreement on a realistic deadline. Often we don’t need to get into the weeds of what Bob is doing Tuesday afternoon at 3pm, we don’t need to care about that, after all Bob is a well-paid professional and can manage his own time (hopefully).
4) How confident are you in the estimate?
Once you have discussed the total work hours and therefore resulting elapsed time think about how confident you are in this estimate. Is it based on past experience or a finger in the air? It’s worth asking how the team member came up with the work hours estimate, how much science is there in their thinking? The ideal is to refer back to historical data. “We did this task last year and the project schedule shows it took 3 weeks” is a great place to be, but so often this historical information is not available.
As a minimum you might tag each task with a “confidence factor” of low, medium or high. Those tasks highlighted as low confidence maybe worth further investigation or at least greater attention once they are underway.
5) Review progress against these estimates
Once each task is underway, monitoring progress against original estimates is very important. How are we progressing compared to our original estimates? Actuals data not only informs us about what has been completed but should also be used to help refine future projections. If everything so far has taken 10% longer than originally thought, perhaps this same variation should be applied to forecast work.
The very best estimates are based on good data. The information we have, particular from previous projects, then the better our future estimating can be. However, is your organisation setup to capture this information? Is anyone taking data from completing projects and presenting it at the start of new projects? This is one of the benefits of a good Project Management Office (PMO). Food for thought for your organisation.