Keys for Creating an Accurate Project Time and Cost Estimate
Figure #1: Sometimes, You Have to Fight for the Truth
Why Accuracy is Important
As project managers, we are often pressured to create estimates very quickly and with too little information. Worse, we are pressured to say what the executive sponsors want to hear: "I only have $15,000 for this database upgrade. You can do it, right? And it's got to be done in two months, too!"
Yet, if we say yes, and it is just not possible, that makes things worse. We don't look very good two months and fifteen thousand dollars later, when we come back and ask for another ten thousand dollars, and another month. And yet, 3 months and $25,000 was the best anyone could do, and we could have known that up front.
Unfortunately, many organizations have gotten used to accepting low estimates, and figuring that most projects will go over budget and be delivered late. According to the Project Management Institute, 55% of all projects do not deliver satisfactory results. The percentage of projects that deliver satisfactory results on time and under budget is way below that.
Project managers, too, have come to believe that we can't make realistic estimates, or, even if we could, we can't do it within the time available to us. And, if we could, the boss wouldn't want to hear it.
To put it simply, the professional world has come to expect bad estimates.
That means we'll look really good if we can do good estimates. Now, I'll be honest, presenting them won't be easy. Convincing our bosses that we need good information and enough time to estimate well will be a challenge. (See the raccoons in Figure #1.) Delivering realistic estimates will mean, at first, presenting higher project budgets and longer schedules - more arguments. But look what happens on the flip side: We start to deliver our projects on time and within budget - or darned close. And we're the only ones doing it.
That's worth it's weight in gold.
In estimation, bias is a pressure that leads us to come up with an estimate that is either too low, or too high. So the first thing we have to do is to want to make an accurate estimate. Just remember that, at the end of the project, you want to look back and say: My estimate was right on; my management was good; and I delivered right on target.
In fact, one way of understanding bias is to think that it is something that makes you consistently off target. In archery, a cross wind will blow the arrows off target in the direction of the wind. That's bias. We compensate for bias by being aware of the pressure and refusing to let us throw us off.
Here are some things that could throw us off, and what to do about each one:
- Required or imposed budgets, and statements of available funds. A required or imposed budget is a statement of what the project sponsors, or the customer, is willing to spend. A statement of available funds says how much money is in the budget. Be clear that none of these has anything to do with an estimate. Either don't know that amount before you estimate (like the blind lady justice in Figure #2), or, if you know it, set it aside. After you make your realistic estimate, you can present a comparison of the estimated budget and the required budget.
- Wishful or resentful thinking. If we like our boss, we will want to fit the job into what he says he can afford. If we think our boss is unrealistic, we will want to demand twice as much. These are natural, human reactions of friendliness or resentment. Please set them aside. Focus on the professional pride that comes with creating a realistic, accurate estimate of work to be done and its cost and time.
- Wanting to just use a copy of an old project estimate. Do not simply copy an estimate whole from another project. There will be two errors if you do. First, that estimate may not have been accurate. Always use accurate actual cost statements and schedules, not past estimates. Second, no two projects are alike. Even if your project is very similar to a past project, focus on the differences, not the similarities.
- Estimating the entire project as just one thing. This doesn't work. Accurate estimates are built on details. Think of how much each piece of the project will cost, and how long it will take. You will get a more accurate picture. Also, if one item is high, and another is low, it will all balance out.
Now you know how to avoid the pressures that create bias. Let's look at some tips for making a great estimate.
Tips for Making a Great Estimate
Here are some things you can do to make a great estimate:
- Make sure that you know what the project is supposed to deliver. Without a clearly defined scope, an accurate estimate is impossible.
- Give yourself as much time as you can. Your first few estimates with a new method will go slowly. After that, you'll know how long you need.
- Get good information. Don't just buzz off a few emails. Make time to talk to people, to explain what you're doing, to make sure they understand, and to make sure they think about what you are asking and give you good information.
- Focus on the unique pieces first. Every project has some unique elements. Estimate those carefully first. They you can copy from past project actual budget and schedules for the parts that are more routine.
- Don't leave anything out. Once bias is eliminated, the biggest estimation errors arise from simply forgetting a piece of the project plan.
Estimation Process and Techniques
In a medium-sized or large project, we may create two estimates. The first, or early estimate, is done after we have a clear scope statement, but before it is worked out exactly in the Work Breakdown Structure. The early estimate can be created by comparison to early projects, by the Bucket Method, or by a sophisticated (and expensive) technique called Modified Wide-Band Delphi.
After the WBS and activity list are complete, very accurate estimates are created using PERT, the Project Evaluation and Review Technique.The relationship between the project planning process and early and late estimation is described more fully in Project Management Time & Cost Estimation Techniques: An Overview.
Wherever you are in the project planning process, and whichever estimation technique that you use, all of the methods in this article to prevent bias and make estimation easy apply.
Making a Good Estimate - Fast
As you make your first few estimates, you'll want to set things up to do it faster the next time. Here's how:
- Organize and store your data. You'll be using actual data from past projects (schedules and budgets) often. Make sure it is readily available, well-organized, and, if possible, in standard formats and searchable. Ideally, it should be in a format that is easy to copy from, such as a database, a spreadsheet, or an editable document, rather than a print-ready document such as a PDF file.
- Make a contact list. You will find certain people are really helpful. They understand the type of data that you want. Some will be other project managers. Some will be accountants. Some will be vendors who are happy to do a quick estimate, in the hope you will buy from them. Make a contact list of people who are especially good at giving you information with estimates, and note their areas of expertise.
- Build templates. Build your own estimation templates, forms, and spreadsheets. This will make things really fast the next time. Be sure you clear all the data from the current estimate from your template - you don't want errors to creep in. Having these templates also makes it really easy to train others to help you get your estimates done faster.
- Build a team. The fastest way to check numbers is to have one person read the source, while the other checks the final document. You can also split up parts of the project, and have workers estimate their own time and costs. This is far more accurate, and also encourages to own the estimate, so that they try to live up to what they said they could do.
Lastly, when the project is done, compare what actually happened on the project to your estimate. If your estimate was right on, you've got a great process. If your estimate was off - and that's okay, sometimes it will be - then dig in and find out why, and update your templates to improve your process.
You'll be a fast, accurate estimator in no time. Well, no, that's not accurate. That's biased hype. You'll be a fast, accurate estimator very soon, and sooner if you give it your best effort!