Costs are one of the main concerns (if not the main concern) of decision makers in project management. How much will this project cost me or my company? Can we afford it? Is this worth it? How much return will it bring to my investment?
These and thousands of other questions populate the mind of your investors when they are considering whether to approve or not your project. This is why it is extremely important to present an accurate estimation of your schedule and costs at each phase of your project.
In this post, we will discuss the fundamentals of accurate estimations, several tools that will help you and basic principles that should be observed in every cost estimation procedure. So let’s start!
The Basics of Estimation
The more unique a project is, the harder it is to produce accurate estimations. Among the factors that increase the complexity of estimation, the most common are:
- The project manager might not be familiar with everyone in the project team. As the complexity of a task increases, more productivity and specialized skills will be demanded, making estimation of time and budget more difficult.
- New technologies are not always reliable and they might present some setbacks while being implemented or used during the project.
- An imprecise schedule with mistakes in the timing of tasks will require more budget for the project. If, for example, your project is behind schedule and you are to receive soon a critical part that cannot be stored for a long time, you might have to increase productivity (therefore, costs) or pay the supplier an extra fee to return in a posterior date.
In the process of estimating the budget and the schedule of a project, you might run into some well-known mistakes. This happens not because you cannot carry an estimation properly but because it is very easy for everyone to fall into these mistakes due to lack of attention or need for urgency.
The Most Common Mistakes in Estimation
Although common sense and a bit more complex thinking would suggest you to avoid these mistakes, it is very likely that you will (or already did) make them in the course of estimating a project.
Guessing Estimations before Actually Computing Them
You enter the lift, the door closes, and you look around. You notice that the person standing next to you is a high-level manager who just happens to be very interested in how your project is going. “Hey, how about that cost estimations you’ve been calculating? Anything yet?” the manager asks. In the pressure of the moment, you inevitably provide a very optimistic guess with very little empirical ground. You simply forget that advice your subordinate gave you about the probable increase in costs across your supply chain. What happens here is that you just ignore the “worst possible scenario” and jump right into the “best possible scenario” to shout some numbers that will please the high-level management. “Oh, very well! Nice figures!” she says while leaving the lift… Until she finds out those were very optimistic, completely unattainable numbers. So how to deal with it? How to avoid such optimistic guessing? Here I present a few tips to help you:
- Make a concerned expression (or at least pretend you are concerned) and explain that a lot of factors are influencing the efforts involved in the project. Be honest and say that you still don’t have all the information to offer a final number for the estimations.
- Write down what the manager is asking, as well as the questions you need to address before giving a relevant estimation. This is important for two reasons: (1) you avoid giving an estimation too broad and (2) you check whether the manager has a clear idea of what he or she is looking for and whether he or she can answer your questions about the estimations.
- If, after all that, you still got this annoying manager breathing down your neck, you have three options: (1) You can give your most precise number and list down everything that could go wrong. (2) You can take your most precise number and double it. And double it again. And then say the result to the manager – I particularly like this one, as it addresses the irrationality of the manager with irrationality in the estimation. Just be careful so the manager doesn’t think you are playing around with him or her. Finally, (3) you can simply refuse to give an estimation until you have better information and more precise numbers.
Starting an Estimation before You Have Thorough Specifications
Who of you tried to estimate the time required to study one chapter of a book and, in the end, took at least two or three times more than planned? This happens because of many factors (including your occasional “checking the Facebook for 5 minutes” break), but the lack of complete specification is a very relevant one. The recommended process would be to think about how much time we actually need to go through a page – how much time needed to read, to make notes and to solve the problems – and then multiply this time by the amount of pages in the chapter. But who actually does it? What we usually do is to take the entire number of pages, say 50, and think “hmmm… 3 hours sounds good for 50 pages!” However, if we think rationally we could expect something around 15 minutes per page including reading, notes and exercises (doesn’t sound much, does it?), so the total estimated time would be more around 12.5 hours – more than four times our initial estimation!
Ok, maybe studying is not your favorite activity, so let’s think about other examples. Think about a construction company: would you expect it to give a precise cost estimation if it doesn’t have access to a blueprint? What about raw material and labor hours required? All this information should be precisely described before you start your estimation. This is why we build a Work Breakdown Structure!
Despite being obvious, this mistake is more common than you might expect. How many of us start doing something because we think it is a “good idea” but stop pretty soon due to “higher effort, costs and time than expected”? I myself started tons of blogs, always promising that “this will be the one”, and after two or three weeks, I would never touch it again. Thank Google and WordPress for free hosting, otherwise I would have wasted a considerable amount of money.
Padding the Estimate
There is nothing wrong with accounting for uncertainty in an estimation. Maybe you have real reasons to add more time or budget to your estimates because you think an adversity may happen with probability significantly larger than zero in the future. The problem is when you add more time and money on purpose just to deliver before the biased deadline or under the biased budget. There are several problems that may emerge. If you pump up the numbers, it might be the case that your project will be perceived as a bad investment and will not accepted. Another drawback involves your reputation and the trust of your managers on your work. Do it once (and get caught) and you will always be seen as someone unreliable.
There are two alternatives: (1) make sure you don’t get caught – but this is out of the scope of this course – or (2) offer honest, detailed estimates. Everyone knows it is impossible to predict the future, so offering your best shot will give you some reputation to fight for a higher budget or a longer schedule when necessary.
The Golden Rules of Estimation
Now that we saw what not to do, let’s move to the good part: universal rules applicable to all projects at all times.
Choose the Right People
Don’t take the new interns to do your estimate. They might not be familiar with the job and are certainly not familiar with the operating system of your company. To have the right people do your estimates means three things:
- Bring experienced people on board: understanding how the work is done is a prerequisite for good estimates.
- Bring the people who will perform the work: while the top management will tell you how much the productivity rate should be, people doing the work will tell you how much the productivity rate can be (assuming they are honest workers).
- Bring people that know the science and art of estimation: while knowing how to do the work is important, someone should not be allowed to produce an estimation if this person does not know how to make one. Estimation is not simply forecasting the most optimistic scenario; it is applying judgment-free rules to obtain the most accurate estimate of real-life scenarios in the future.
Use Your Previous Experience
A construction company creating a new estimation for a building should incorporate all the knowledge it has collected from previous estimations. If the suppliers are likely to delay the delivery of raw material by two weeks, it is recommended to include this in future forecasts. If you know your historical absence rate, include this in the estimations. Since more historical data reduces the need for guesses and gut feelings, the more information you have the better.
Focus on the Equilibrium rather than on the Estimate
It is very easy for someone to crack under the pressure for lower costs or tighter schedule by higher managers. If you built your estimates following the estimation principles, there is nothing to be afraid of. It doesn’t really matter if your boss comes to you and say “you should reduce these numbers, the budget is too high!” Your answer should be something along the lines “(dear sir manager, ) These estimations were built based on the historical data available to us at the moment. Considering the features of the product, this is a realistic estimate of how much time we will need and how much money will be invested during the process. It is not possible to reduce the numbers without affecting the reliability of the estimations. If you want to change them, we should focus on changing the product or the processes themselves.” In other words: you shouldn’t change the numbers because of pressure if they are based on unbiased estimation processes. If your manager still wants to do it, you should make it clear that he or she should be responsible for the changes (and the potential consequences, good or bad).
Focusing on the Optimal Degree of Accuracy
While more accurate estimates are always a good thing, increasing the preciseness of your numbers costs money. Therefore, it makes sense to provide different degrees of accuracy at different stages of a project proposal or planning.
The first level of accuracy is the so called ballpark estimate. While severely biased, this estimate doesn’t require much of your time nor resources and it relies on the experience of the person providing it. The only goal of this guess is to check whether it is worth to proceed with more detailed estimations for the project.
The next level, still considerably variable, is known as Order of Magnitude or Rough Order of Magnitude. It is calculated based on the information from previous projects, and it tries to find similarities or relations between past projects and the current estimation. For example, if our construction company will build a house three times as big as a previous construction, it makes sense to give three times the cost as an Order of Magnitude estimate.
The third level, the detailed estimates, are where you should spend most of your time. Notice that we don’t reach the third level without getting an approval in the first two. Still, the detailed estimates, or the bottom-up estimates, are the phase where you will provide a serious estimation for the project. The detailed estimates are usually calculated using the bottom-up approach: you start by estimating the resources involved in different phases of your project and you bring everything together under one final calculation. The bottom-up estimates bring information about schedule and resources, as well as the forecasts of budget and cash flows.
Techniques for Accurate Estimations
Ok, enough chit-chat. We know what we should do and what we should avoid, but we don’t know how to do it. Let’s discuss several techniques for creating estimations for your projects. Naturally, a single article cannot cover all of them in their full extent (there are entire books written about estimation, so…), but I can give you a good idea about how each of them work. We will review the basic building blocks of estimation, so if you understand what we will discuss here, you will be able to understand any estimation procedure and software.
Phased estimating is about estimating the different phases of a project. As a consequence, it requires information about one phase at a time. This is very reasonable, since requiring a single number estimate for an entire project before we start is, most of the times, implausible. Let’s see this in a graph to understand how phasing estimating can improve the decisions about a project.
In practice, we should always have two estimates: a detailed estimation for the next phase of the project and the Order of Magnitude for the rest of the project. Another important aspect that the phasing estimations recognizes is the high uncertainty of a project in its initial stages. It deals with the uncertainty by breaking down the project into several phases, each of them considered a project itself.
As you can see in the graph, the initial phase starts with an Order of Magnitude estimation for the entire project and a detailed estimate for the first phase. Once this is over, the process of estimating starts again for the next stage, and a new authorization is required. Also, a new Order of Magnitude – which will be much more precise than the first one due to the learning effect – is calculated for the remainder of the project. This cycle of adjusting the Order of Magnitude estimate and producing detailed estimations will repeat itself for each new phase of the project.
While the phased estimating is appreciated by the project staff – as it requires focusing on one phase at a time -, sometimes the sources of funding do not like the fact that they don’t have an overall detailed estimation of how much everything is going to cost altogether. You can solve this problem by explaining that an overall estimate without enough information is more likely to be inaccurate and put everyone at a higher risk. The big advantage of the phased estimating is that it allows the project to be planned and executed through many small, informed decisions rather than by one large, uninformed decision.
Apportioning, also known as top-down estimating, starts from the top with a total figure for the project and then assigns percentages of that budget to different phases and tasks involved in the project. Again, the Work Breakdown Structure is essential for you to have a clear picture of which tasks will be involved in the project, as well as their complexity. As the top-down estimate derives the smaller budgets from the top, we must be careful to provide a good number at the beginning. If the overall estimate is not valid, the smaller shares of the budget will also deviate from reality. Let’s see how the apportioning estimation looks like.
While the bottom-up estimate is usually preferred since it is more accurate (we will see it soon), apportioning is easier to implement and can be considered appropriate for selecting which projects your company should invest on. Apportioning can also be combined with phased estimation to allow for updates in the percentages for each task as the budget becomes more and more realistic through the course of a project. Let’s see a practical example. Suppose that our apportioning division from before is now attributed to three departments: Research, Design and Implementation. Have a look at the chart below.
Let’s understand how we can use the information from previous phases while creating the Order of Magnitude estimation for the next stages of the project. First, note that our initial guess was $300,000.00 to complete the project. As Research is supposed to consume 30% of the budget, the initial estimation for Research was $90,000.00. Once the Research phase is completed, we should proceed to updating our numbers. We do that by incorporating historical data into our model: the actual cost of the Research phase of the project. This cost was only $72,000.00, less than our initial estimate. Since this cost is supposed to account for 30% of the entire budget, we can update our estimations of the costs of the project as follows:
The new total cost of the project is $240,000.00, and the revised cost of completion is simply the updated total cost minus the already spent Research amount ($240,000.00 - $72,000.00 = $168,000.00). Finally, the reviewed estimate for the Design phase is 50% of the new total cost of the project: $240,000.00 0.5 =$120,000.00 .
The parametric technique seeks to identify parameters (or factors) that play an important role in defining the estimation of a project. Consider the following example:
A construction company was hired to build a house of 150 square meters and is now working on the schedule estimation for the project. From previous construction works, it knows that a house of 100 square meters is normally built in 180 days. Therefore, a first estimation for the house would be to take 270 days to be built. However, what the company also observed is that:
- Building the foundation requires twice as much time as building the house;
- The required foundation for the 150 square meters house is twice bigger than the foundation for the 100 square meters house.
- The weather in the region of the new project is less favorable. In order to incorporate this in the model, the company decides to add 20% for weather delays;
- The supplier of raw materials is closer to the new project than to the old one. This allows the company to be more optimistic and reduce the overall time estimate by 5%.
So how would our new estimation look like?
First we need to figure out how time was divided in the first house. To make things simple, let’s assume that building a house has only the foundation and the construction phases. Based on that, we can figure out how much time we used for foundation and for construction in the first house:
So the time spent in the foundation was 120 days, while the time spent in the construction was 60 days. We also know that the new house requires twice as much foundation, so it makes sense to estimate the required time as being 240 days. By using the same proportions as before, the estimation for the construction of the house is 120 days.
We are now dealing with an estimate of 360 days instead of our initial 270. Finally, we have to make the adjustments for weather and location of suppliers:
So our final estimate obtained through parametric estimation is approximately 411 days, 141 (or 52%) more than our initial figure. Pretty big difference, huh?
The relevance of parametric estimation is that it allows you to incorporate past data into your estimates. As you can see, we included the information that foundation requires more work, as well as supplier and weather conditions.
Consider another example of the remodeling of a house. The company is now estimating the cost necessary to paint the inner walls of a rectangular saloon and its windows. From market research and previous experience, it came up with the following numbers:
- Cost of 5 liters of paint gallon is U$ 80.00, or U$ 16.00 per liter;
- 5 liters of paint are just enough to paint 40 square meters of surface, meaning that 1 square meter requires 0.125 liters of paint;
- A normal worker can paint 30 square meters per day at a cost of U$ 15.00 per hour, or U$ 4.00 per square meter (assuming an 8 hours working day).
The surface of the saloon is then given by2h(l1+l2), where l1 is the length of the shorter wall and l2is the length of the longer wall. We can then come up with two parametric equations to estimate the cost and the time necessary to accomplish the project. Notice that our parameters are (1) the height h, (2) the lengths of the walls l1 and l2, and (3) the number of workers in the project.
If we want, we can include percentual margins to account for waste of paint, absences from workers, etc. These would simply require us to adjust our previous estimates to include the margins.
Again, the estimations are based on previous information from the company’s experience. In addition to that, the model we derived above is valid not only for the current project, but it can be adapted, with minor changes, to more general equations that will give estimations of costs and schedule for more complex projects.
Although the name sounds complicated, parametric estimations do not need to be so. The important thing is to use your historical data systematically. Parametric estimation is usually more precise for tasks that are lower in the Work Breakdown Structure and it can be combined with phased estimating to produce accurate estimations for each phase of the project. Finally, the more detailed the specification of the product or service being executed in the project, the better your estimation will be.
Bottom-up estimation, as the name suggests, starts with estimations for the more detailed tasks and walks upwards in the Work Breakdown Structure to produce the final estimates of each summary task. We already did a big part of the bottom-up estimating process when writing down the Work Breakdown Structure of our project. Remember we also discussed the estimation of the required work for each work package and how we brought them together in the project plan. As our example was not very detailed (in order not to make the article about schedule estimation tediously long), I will provide a complete and sophisticated example in the eBook about cost and schedule estimation. You will be able to see how each of the stages of the planning process work in practice, as well as how resource leveling and other tools can be used to better estimate costs and schedules. For now, let’s check how bottom-up estimating looks like.
As you can see, it starts with the lower tasks and moves up to provide aggregate estimates for summary tasks and for the final level, the project itself. Ideally, bottom-up would always be used, but that is just not possible. If we are at the beginning of a project, we simply don’t have enough information to estimate every single work package precisely for the entire project. The best way to use bottom-up estimation is to combine it with phased estimation to produce a detailed analysis of each stage of the project. You can also combine it with parametric estimation and bring your predicted figures to an even higher degree of accuracy.
Creating a Detailed Budget Estimate
Ok, so you got your initial estimation and, luckily, the board of directors said that your project is nice and should be taken forward. Now the real work starts. Think about applying to a university or so… You send your past information and motivation letter, but it is only after you get selected that you will start actually collecting all the necessary documents and make the arrangements to start your studies. The same thing happens with projects: you get your initial estimates, build a nice document and submit for approval. But it is only after you receive the formal authorization that you actually start collecting detailed data and producing detailed estimates of your tasks.
Defining the Sources of Data for Your Budget Estimation
While calculating the budget doesn’t require many more skills than simple mathematical education for adding up some numbers, the trouble lies on how to obtain the data you will multiply and add. Which categories should be included in the budget estimation? Which costs should be considered? Let me present you several major categories that almost every project should focus on (it might be the case yours will have different ones you should also include, but this varies from business to business).
Internal Labor Cost
The internal labor cost refers to the amount of labor required from people employed by your company. In the estimation of the schedule we already discussed the basis of how to include the internal labor costs, but in the cost and schedule estimation eBook I will present a complete example of how to estimate your costs of labor in a real world scenario.
An accurate amount of labor comes from having accurate amounts for the work packages in your breakdown structure. Remember what we discussed about time constraints and resource leveling? These practices provide a very good estimations about the amount of work required for each stage of the project, and what is left is just to multiply it by the hourly rate (or the rate for the relevant period of time being used in the analysis) in order to get a final number for the budget.
Since the hourly salary of your employees might differ among each other, the best way to come up with realistic numbers is to use the burdened rate. The burdened rate is nothing more than the average total cost of an employee to the company. Using just the hourly payment of each employee is not good enough because people cost more than just the salaries they receive. Often there are benefits they are paid, absences we have to account to, overhead costs (such as fixed costs mitigated among several projects which are not from one specific project, so they can’t be assigned to it) and so on. So how to get this rate? It couldn’t be simpler: go to your financial department and ask. Almost every company has this information, so you really don’t need to spend efforts trying to figure it out.
Internal Equipment Cost
The equipment we are talking here is not the routine equipment your workers are assumed to have. Of course everybody in a software company will have a computer; of course people working on a construction will have access to trucks. The equipments we are talking here are those which are not normally available. For example, say that our earlier project to build a house with 150 square meters requires a special bulldozer because of the need for a stronger and deeper foundation. The cost of renting this bulldozer should be included in your project since it is not an equipment your company own and it will be rented specifically for this project.
If an equipment will be used in a single project (for example, the rented bulldozer), all you have to do is include the costs of this equipment in the project estimations.
Now consider you are thinking about buying the bulldozer instead of renting it, since you have a positive forecast of other projects which will require this equipment. How should you include it in the budget? It doesn’t really make sense to include all its costs under one single cost estimation, as this will drastically increase the costs of your project and your customers will certainly raise their voices. In order to overcome this problem, the best way is to estimate the total life cycle of the equipment and to divide the initial cost in smaller units. You can, for example, estimate that the bulldozer will be useful for 50,000 hours of work, so a simple approach would be to divide the cost of the bulldozer by 50,000 and multiply the cost per unit of time by the estimated amount of hours the machine will be employed in a specific project. Although this estimates based on estimations, it is a good way to avoid attributing all costs of equipments to a single point in time.
Costs of External Labor and Equipment Use
It might be the case that you will hire someone to do some parts of your project. How to include the cost estimates for this will depend on what your contract says. If it determines a rate used by the vendor to bill labor, equipment and materials, you will use the rates and the estimated amount of work to calculate the estimated cost of that service for your company. If, however, your contract states a fixed fee, the work of estimating everything will be done by the vendor and you will just need to worry about including a single final number in your project. Of course, you should keep an eye open in order to make sure that the estimates are accurate and you are not being overbilled.
The cost of materials is one of the few components of the budget that is not calculated solely based on the Work Breakdown Structure. The product or service specification is the best place to look for information to estimate the cost of materials, although the Work Breakdown Structure can be of some help.
Generating the Schedule for the Cash Flow of the Project
Now we reach the point where you know everything about how much you plan to spend during the execution of the project. Very good, but we are still not done. In addition to how much you will spend, the information about when the money will leave your pocket is almost as relevant as the total amount.
Once you have the information about your schedule and costs, timing the cash flows of your project is no big deal. It is just a matter of identifying the key points of your schedule and link them to the estimated amount of money to be paid at that time.
What Is Coming in the Next Article
Ok, I know… Like me, you probably want some action now. You have a solid knowledge of the basis of scheduling and estimating, but a complete, sophisticated example is still missing. The next article will fix that. I am preparing a more complex example that will cover everything from building your Work Breakdown Structure until estimating your costs and schedule precisely.