As I alluded to in my post last week, there are many ways to perform a cost estimate – all are right, but the key is for your organization to pick the approved standard, so that confusion can be minimized as much as possible. The standard should include often-overlooking details, such as what unit of measure the cost estimates will be done in (Dollars? Euros? Etc.), what level of precision the estimates should be rounded up or down to (whole dollars? Thousands?), and what the acceptable level of accuracy is (+ 10%? + 20%?) to be used. Cost estimate accuracy almost always improves as more details of the project are known, and this should be understood regardless of the project environment or method being applied.
Cost Estimation Methods
Analogous Estimates – This type of estimate depends on historical estimates of similar projects, using elements of scope, budget, duration, complexity, and others. While this method can be faster than other methods, it is also typically less accurate and, when used, should account for the Net Present Value of the historical costs to be used.
Bottom-Up Estimates – This type of estimate typically estimates each work package (or individual activity within the work package), and generally provides the greatest level of detail and accuracy (+ 5%), which is then rolled up into a summary for each higher level component to achieve an overall project cost. These detailed estimates typically come straight from vendor quotes, unit prices, and fairly complete plans.
3-Point Estimates – Based on the Program Evaluation Review Technique (PERT), this 3-point estimate gives the most likely estimate 4 times more weight than the pessimistic and optimistic estimates. This estimate is especially beneficial when estimate unique types of projects that have little to no historically similar project data to use – such as innovative efforts in the research and development arena. To calculate the 3-point estimate, you simply determine the most optimistic (O) estimate, the most pessimistic (P) estimate, and the most likely (M) estimate, and calculate it as follows:
3-Point Estimate = (O + 4M + P) / 6
Order-of-magnitude estimate – This type of estimate may use historical data, but the projects don’t have to be similar (as they do when doing analogous estimates), and are not usually very accurate (+ 35%). These are top-down estimates typically applied to a level 1 of the Work Breakdown Structure (WBS) and excludes detailed data, although parametric estimates (based on statistical data) may be included in some cases.
Learning curve estimate – While this type of estimating stems from the manufacturing industry, it may be applicable to any industry or project that requires repetitive functions where continuous operation (learning) will lead to a reduction in time, personnel, or money it takes to do the repetitive function.
Top-Down estimate – This type of estimate has no detailed data for the basis, but is prorated from previous (similar in scope and capacity) projects. For example, you may say that this project is 25% more difficult than the referenced (previous) project activity, therefore, requires 25% more time, labor hours, material, dollars, etc.
Contingency and Management Reserves
One of the biggest pet peeves in project estimating is “padding the estimate” – when the actual estimates are on-purpose padded for “just in case”. Why is this a pet-peeve? Because when this is done, the project costs are almost always over-estimated and the actual spending isn’t tracked well (because nobody wants to “admit” to padding their estimated costs, and it enables sloppy monitoring and control, and skews historical use of the data for future projects, as in the analogous method). Instead, it should be understood that all costs estimates are just that – ESTIMATES. They will not be exact, and they aren’t supposed to be – the “actuals” are the exact costs.
Where should this “padding” take place, then? Instead of simply increasing cost estimates within given work packages, the more appropriate approach is to include a contingency estimate that is meant to cover the “known-unknowns”. If you know you are embarking on a high-risk project, you’ll want to include a higher contingency estimate than if you were taking on a project that is similar to something done before, or with low risk. My use of contingency reserves most often were done as a percentage of total project cost (such as 10% for high risk projects), but you can also estimate contingency costs as a fixed number, and you don’t have to base it on the total project – you may choose to estimate contingency only for specific activities. The key is to call out your contingency estimate so it is included in your total project budget instead of “padding” other areas of the project, essentially hiding it, in an attempt to achieve the same thing (or to cover up for bad planning!).
Unlike contingency estimates, estimates can also be produced to cover management reserves, but these estimates are intended for management control in covering the “unknown unknowns”. The management reserve estimate is actually NOT part of cost baseline, but IS part of the project budget. Whenever management reserve funds are used, that amount is then added to the project cost baseline and must be an approved baseline change.
Bottom line? Use the management reserve to cover unforeseen problems that occur on the project, as long as they remain within the scope of the project. Don’t use the management reserve to cover up bad planning estimates or budget overruns (nobody does this, right?!).
Which estimating approach does your organization use? Is it something other than those listed in this article? Stay tuned for access to my free cost estimating templates, or sign-up here to be notified when you have access to them (and can download and put to use immediately)!