One of the biggest challenges for project managers is working with bad estimates.?
I was once responsible for the estimating model (EM)? for a multi-million dollar consulting practice. Here are some good estimation practices I picked up:?
- Investing in an estimation MODEL is a valuable exercise for any team that does a high-volume of project work. I’m grateful to a colleague of mine (
Loren Davie
) who taught me about an “estimate primitive” which can represent a fundamental, scalable architecture (defining factors such as activities and risk, and their relationships to each other). Resist the urge to build custom factors for every new activity type.?
- There's a lot of potential for advanced tooling to help with estimation and operations, but there needs to be stakeholder buy-in to maintain them.
- It’s hard to build a good EM without a clear strategy and design for your services portfolio. Know what you do, what you might do, and what you don’t do before designing our EM architecture.?
- An EM helps to build custom projects, but it also can help with designing services, bundling services into standard offerings, and strategic planning.
- Using an EM as part of a project scoping and negotiation discussion helps keep commitments realistic during pricing and timeline negotiations. A good EM can also reveal alternative options for how to problem-solve.?
- The most accurate estimates are made when both an outside expert and an accountable member of the delivery team participate in the scoping discussion.
- Both fixed scope and fixed hours projects can use an EM to manage expectations and prioritize work.
- Tracking hours lets you go back and learn about estimate accuracy and make adjustments to the model and train estimators. If an EM can pre-load into the project management software it can also save the PM time and help drive budget discipline.?
- Creating adjustable factors for risk or uncertainty create a cushion across a portfolio. Within a project, these factors can apply to individual tasks, workstreams, or to the project as a whole. It’s helpful to have a set of evaluation questions to inform how to set these factors.?
- Pricing and estimated effort don’t need to be directly related. If you price to value and willingness to pay, you may be able to achieve much better margins as you build efficiency and predictability into your service design and delivery. You also may choose to price below the cost to serve if you are creating new services or investing in the customer relationship.?
One of the best stories I know about estimation error is from Daniel Kahneman’s Thinking Fast and Slow, regarding a team’s estimate of how long it would take to write a book (p245-247). The team thought it would only take two years - but it actually took eight. This story highlighted three fallacies we have to watch for in any estimate: inside view (assuming things would continue as they had already), planning (assuming best case), and irrational perseverance (continuing when actual evidence suggested a negative ROI).??
Whether it’s creative or construction, selling or sailing, discipline in the estimation process is a chance to avoid predictable errors up front, and learn from unavoidable errors after the fact.?
