Join Bonnie Biafore for an in-depth discussion in this video Estimating time and cost, part of Project Management Foundations: Schedules.
When you start planning a project, you need to estimate what it's going to take. Estimating can be tough, but there are ways to make it easier. The first step is to get everyone to agree on the level of accuracy for estimates. There are three levels of accuracy to choose from. You can start with a rough estimate, plus or minus 50%, to see if a project makes sense to pursue. As you get additional details, you can put together more accurate estimates. The second level of estimate is the mid-range estimate. These are usually plus or minus 25%.
When you finish one project phase, or a pilot project, you can re-estimate the next phase based on what you've learned. The third level is the detailed estimate. These take the longest to develop but they should be really close to your final numbers, plus or minus 10%. The second step is to estimate tasks. There are many ways to do this. If you have information about a similar project in the past, or team member's performance on other projects, by all means, put that data to use. Estimating based on previous actual results is a great way to get an accurate estimate in no time.
Another way to obtain accurate estimates is to get people who understand the work to estimate it. For example, if you know your team members, give them the big picture of the project and describe the tasks you need estimated. Have them estimate how long it would take them to complete the work. Another option is to track down experts, like senior people or external consultants and ask them for estimates. There are two sensible actions to take when you get estimates from others. One, you can trust that the estimates are good or two if you have your doubts, go get a second opinion.
Expanding on the concept of getting a second opinion is an estimating technique called the Delphi method. It's based on the idea that more heads are better than one. You can ask several experts to estimate independently. You share the results with the group and ask them to re-estimate. Repeat this step of sharing and re-estimating a couple more times, and then use the average of the last round as your final estimated value. Sometimes people share the reasoning the experts give for their estimates, others simply share the estimate values. Interestingly, the Delphi method seems to produce good estimates either way, so you can use whichever approach makes sense to you.
Here's an example of applying the Delphi method. In the first round, one estimator comes in on the high end with 200 hours. Another estimator is on the low end with 150. And the third estimator is almost in the middle with 180 hours. In the second round, the high estimate comes down to 190. The other two increase their estimates. Maybe because they thought of some work they forgot. The third round high and low estimates continue to move toward the middle. While estimator three sticks with her second round number. The estimates in the fourth round seem to be stabilizing, so we'll take the average of this round as the estimate.
Add the three estimates together and divide by three to get 180 hours. Another way to get more accurate estimates, is to take an iterative approach to the project work. You start with a small piece of the project, estimate it. Deliver the work, and see how long it took. That way you can come up with better estimates on the next portion of the project. Another approach is PERT, which stands for program evaluation and review technique. You ask people to produce optimistic, pessimistic and most likely estimates.
For example, the optimistic estimate for a task is 16 hours. The pessimistic estimate is 32 hours. The most likely estimate is 20 hours. The weighted average of these estimates comes out to be 21 hours. However the PERT method also uses these estimates to run multiple simulations of the project to determine possible outcomes. With statistical analysis of these outcomes, you can determine the probability of the project finishing within a specific time frame.
PERT is a good choice when a project is unfamiliar or comes with a lot of uncertainty. It makes people think about risks and it emphasizes the uncertain nature of estimates. These are just a few ways to estimate tasks. There may be other methods specific to your industry. Ask your colleagues for approaches they recommend. Once you decide on an estimating method, your next decision is whether to estimate from the bottom up or the top down. Bottom-up estimating provides more accurate estimates, if your project is broken down.
Let's take a look at how you estimate from the bottom up. You start by estimating the work tasks in your task list. The next step is to add up the estimates for the tasks within a summary task. The total is the estimate for that summary task. Repeat this step for all the lowest level summary tasks. Keep adding estimates together until you reach the top of the project. The total for you highest level summary task provides an estimate for the entire project.
When you estimate from the bottom up, be sure to add in time for communication, management, travel and so on. Estimating from the top down, means that you start by estimating the top level sections of your project. Like each project phase or the sections for each high-level deliverable. Then you work your way down, breaking down your estimates to allocate time and money to the lower-level tasks in each phase or section. Top-down estimating is good for earlier, rough estimates, when you don't have detail about the project or tasks. Top down estimating also works when a project is similar to others you've done. You also need to decide which measure of time you're going to estimate, duration or work hours.
Most people think in terms of duration yeah, that materials going to take three days to write. When you get duration estimates be sure to ask the people who provided the estimates what assumptions they made. Is it three days full time for one person? Or three days with the time that person has available? Or is it three days for the entire team? Instead of estimating duration, estimating work hours gives you a better idea of what the work is going to take. Now a team member could say, that material's going to take 24 hours to write.
Once you know how many people you have to do the work, you can figure out the duration. Three days for one writer, or one day for a team of three writers. Estimates are an important part of a good project plan, because they determine how long the project will take and what it will cost. To give your project a good foundation, pick the estimating method that fits your project.
Lynda.com is a PMI Registered Education Provider. This course qualifies for professional development units (PDUs). To view the activity and PDU details for this course, click here.
The PMI Registered Education Provider logo is a registered mark of the Project Management Institute, Inc.
- Identifying the work that needs to be done
- Adding milestones
- Delaying or overlapping tasks with lag and lead time
- Assigning resources
- Balancing workloads
- Adding buffers and baselines to the schedule
- Uncovering and correcting out schedule problems<br><br>
- The PMI Registered Education Provider logo is a registered mark of the Project Management Institute, Inc.