Join Chris Croft for an in-depth discussion in this video Estimating costs and time, part of Project Management Simplified.
- So now we come to step three, which is the estimating. And we want to estimate for our project what are the tasks going to cost and also how long are they going to take. So we're talking about the money and the time in exactly the same way here. Now, I want to use the example of driving home just to illustrate this. Let's suppose you've got a 30 minute journey home. On a good day if you're really lucky, you might do it in 20. And on a bad day, maybe it'll take you twice as long. It might take as much as 60, okay.
And you may have similar numbers for your project. This might be how many weeks it takes to do your project, and it also might be how many thousand dollars it costs to do your project, okay. So it's exactly the same for the money and the time. The problem we've got is what are we going to tell our customer when they ask how long is the project going to take? Now, if you think about the probability of this, let's suppose that you do a bit of a survey of your journey home, it'd be a bit sad. But suppose you recorded your journey time every time you drive home, and you record them all in a little book and then you maybe make an Excel graph.
What would be the probability distribution of all of your journeys? And it'll probably be something like this, a kind of a skewed normal distribution. So 30 is the commonest, we'll call that the average. And that's the number that most project managers would promise to their customer. They'll say we can do your job in 30 weeks or it's going to cost $30,000. But the problem with this, and you can see on the diagram, is it's a high probability of failure. Approximately 50 percent of your projects are going to come out either late or over budget, 50 percent failure rate.
And we don't want to have a 50 percent failure rate for our projects. Now, you could say, oh, yes, but it's going to average out. So sometimes I've over, sometimes I've underestimated. But the problem is for the individual customer, they don't want to know that, okay, theirs was late but somebody else's was early, they don't cancel out. So the obvious estimate of quoting the average is probably not a good idea. You're going to fail on half your projects. But that's not the worst thing to quote to a customer. The worst thing to quote is this number.
You would be mad to say I can do it in 20. If you quote 20 minutes for your journey home, there's probably a 95 or even 99 percent chance that you're going to be late. So you'd be crazy to forecast that, and yet people do. Why would somebody promise a customer 20? And the answer is they want to make the customer happy. And quite often what happens is that bosses or customers say to you, "If it all goes really well, "how quickly do you think you might "be able to do this job for me?" And like a fool you tell them the truth.
So like a fool you say, "Well, if everything arrives "on time, none of the suppliers let me down, if all the IT "works first time, I might be able to do it in 20." And unfortunately what your customer hears is la-la-la-la-la-la, 20. They don't hear all the clauses you're putting in before that. So the rule is never mention the number 20, once that number's out of the bottle, you'll never get it back in. And if your boss says, "What's the best you could do it in?" You need to weasel out and say, "Well, it's tricky. "We've got some new things we haven't done before.
"It could easily take 40 or 50." Don't say 20, even if you're asked for the best time. And as a boss, it's probably not a sensible question to ask because if you say to a project manager what's the best you might be able to achieve, you don't really want to hear the answer. If they do answer, you don't want to use that number because it's never going to actually happen. So we don't want to be here, and we don't want to be here. Now, of course if would be nice if we could forecast 60 because we can definitely do it within 60. But the problem with forecasting 60 is that nobody's going to believe that, they're going to be thinking, a whole hour to drive that distance? And they're going to look at your project and think, 60 grand for that job, that's ridiculous.
So you're not maybe ought to get away with forecasting 60. And if you're pitching for work, you won't get the contract, so 60 is no good either. So what are we going to choose? And what I would recommend for this, I would recommend going halfway between the average and the worst, which in this case would be 45. Now, that may not sound very scientific, but actually it is. If you're talking about probability here, this is 1 1/2 standard deviations from the mean. And what that means is that only 10 percent of your journeys will be longer than 45, okay.
Only 10 percent of your projects will be late or over budget, you're 90 percent safe if you quote halfway between the average and the worst. Now, that still means you're going to fail on one in 10. You have to live with the fact you will get a 10 percent failure rate. I think most project managers will be delighted if only 10 percent of their projects were late or over budget. And if 90 percent of their projects came in on time, that would be a pretty good success rate. So I would recommend going for halfway between the average and the worst.
Now, you can put in variations on this. If failure is going to be a total disaster, if there's a massive penalty clause or something like that, if it was, let's say, forecasting the number of weeks for the Olympics and you absolutely mustn't open late, then you would push the number out this way. You might forecast 50, 55, or even 60 to be safe. And similarly, if you were pitching for work, you might want to come down a bit lower. But if you do that, you have to be doing that in the knowledge that you're less likely to deliver. So by the time you're promising a customer 30, you're down to 50-50 whether you're going to deliver on time.
And you probably don't want to go that low. So you can have variations on the 45, depending on commercial pressures. But your starting point should be what's true, which is that 45 is 90 percent safe. Most people don't realize this. Most people start with either the average or the average plus 10 percent. So a very common thing to do is just to add 10 percent to the average. But I think you can see on here that adding 10 percent is not enough. 10 percent of 30 minutes is only three minutes, and that's definitely not enough time for your journey home, you could lose three minutes just getting out of the car park.
So don't add 10 percent, go halfway between the average and the worst. And just a final thought about this. Is it dishonest to add a safety margin? So when I tell my customer 45, in full knowledge that the average is only 30, am I being dishonest? And I would say, no, I'm not, the reason I'm adding this contingency isn't really for me, it's for the customer because I want to have a date that's realiable. I want to promise to the customer a date that I can 90 percent be sure that I can fulfill. I don't want to let the customer down.
So I'm adding a safety margin on because I care about the customer. And I'm not going to spend the last 15 weeks of this project swinging in my hammock drinking beer, going ha-ha, the fool, I've convinced him it's 45 when it's only a 30, it's really not that. I know from bitter experience that if there's 30 weeks of work that I know about at the beginning, it's actually going to take me 45. What we're doing is we're adding the knowns and the unknowns, so our planning process is to add up all the known work, which let's say comes to 30 weeks, and then put an estimate on for how much extra unexpected work there is that's likely to crop up.
So what I'd like you to do for your project is just ask yourself, have I got enough contingency in there? Have I forecasted the average or the average plus 10 percent, should I add a bit more contingency in? And if it's already too late for this project, then at least make sure you get it right for your next 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.
- Defining project scope
- Deciding how to list tasks
- Estimating costs and time
- Planning for risk
- Staying on budget