When you are an interactive project manager you are often called to assist in estimating the effort required to bring a project to market. Either you need to do this on your own in areas you have experience in, or you must ask team members and specialists to provide you with estimates.
This is not an easy job – and if done incorrectly could mean lost profitability or overpricing, both of which can be damaging in their own right. So what can be done to improve estimates?
1. Ask for ranges? optimistic/pessimistic/realistic
Ask your subject matter experts to provide you with estimates of their work hours to complete the task. But don’t ask them to only provide a single estimate. Ask them for 3. What is their optimistic guess if they find a few shortcuts or otherwise have no other distractions? Ask them for a pessimistic estimate; what happens if they get hit with some foreseeable setbacks like needing to do research or get interrupted? And then after that ask them to provide what they think will most realistically happen given what they know. What their truest hunch of what will happen is.
2. Do a WBS – break it into chunks
A work breakdown structure is an incredibly useful tool. It revolves around taking your project and breaking it into all of its major tasks. From here you take the major tasks and bust them out into subtasks, and in those subtasks you break them out further into the sub-sub tasks involved there. Once you’ve really broken down the project into smallest units that a single person could do – you can label the whole diagram/hierarchy and seek estimates. There are a few work breakdown structure tools out there you can try (google around) – but doing this on sticky notes or recipe cards can work too to visualize your project’s work units (especially if it is a big project).
3.In a rush/pinch provide estimates with a margin of error (for instance plus or minus 10%)
Sometimes you simply don’t have enough time to get all the information you’d like from all parties. In these cases it is prudent to work through educated guesses, and old projects and ask for quick 5-10 minutes of advice from your team members and cobble this together. It isn’t *WRONG*, because the wrongness is built in – you are sharing the risks of the inaccuracy with your customer (it is a great way to get things done when you have little or no time to operate and a great way to share the risks of inaccuracy upfront with partners/clients)
4. Have more than one opinion on your figures (not just the same developer providing all figures)
Don’t always trust your first estimates (get second opinions!) – especially from people you don’t know that well. Get a few cross-checks with other resources and if possible, people you know well. If estimates have large variances from person to person do a bit more digging to find out why – is it a difference in experience, is it because one of your estimates was badly flawed? Don’t blindly trust that your estimates are always 100% valid on first draft.
5.Get to know how your resources guess
Does Tom always guess high and Stacy always guesses too few hours? Make sure you adjust slightly for known biases
6. Do better detail work on your requirements
Don’t blame the entire estimation process/results on your production staff/resources. It is up to you as PM to define things well. If you have a poor brief, or bad detail and your plan is full of holes and subject to interpretation – your estimates will be BAD. Work on your briefs – get clarification from stakeholders to make your brief coherent. And barring clarity, at least convey any and all assumptions you and your resource team made in order to package up your estimates (this is a good way to manage YOUR risks).
7. CLOSE THE LOOP
Review your actuals vs estimates when the project is complete. This is a classic mistake: not learning from your projects. If you track you and your resources’ time in a time tracker of some sort – go back and see what the actual time spent was relative to the initial estimates. What variance is there? How far off were people? Why was this? What can you learn from this? What are people’s biases or tendencies? If you’re not learning, you’re going to fall into the same traps over and over.
With a little practice you will improve greatly and estimating will seem a whole lot more natural.
How do you handle project estimation for the projects you work on?