Improving The Process Of Cost Estimation

At times, clients have been known to give favored bidders a "last look" at supposedly sealed bids so that the favored bidder can submit a winning bid, often with an unwritten agreement to allow some cost escalation at a later date. There is considerable opportunity for unethical behavior during cost estimation and bidding. Further, estimation and bidding practices vary widely from industry to industry.

Finally, there is plain bad luck. Delays occur for reasons that cannot be predicted. Machinery with the reliability of a railroad spike suddenly breaks down. That which has never failed fails. Every project needs an "allowance for contingencies."

Some writers and instructors differentiate four bases for estimating costs: experience, quantitative (statistical) methods, constraints, and worksheets. They discuss the advantages and disadvantages of each and then, typically, decide that one or another gives the best results. We feel strongly that all four are useful and that no approach to cost estimation should be accepted as the best or rejected out of hand. The best estimators seem to employ an eclectic approach that uses, as one said, "anything that works." The wise PM takes into account as many known influences on the project budget as can be predicted. What cannot be predicted must then, by experience, simply be "allowed for." There are two other factors, particularly common to projects involving intangible outputs such as software programming, that need to be mentioned relating to cost-estimation and the schedule. These two factors have been identified in a classic and highly-readable work—The Mythical Man-Month—by Brooks 17],

First, most projects involve a tangible medium that tends not to be under our control—the wood splits, the paint smears—and thus we blame implementation problems of our "good" ideas on these physical elements. So, when we are working with a purely intellectual medium that has no physical elements, such as computer code, we are highly optimistic and foolishly assume that all will go well. However, when any project consisting of a series of components can only be successful if all of the components are successful, and each component has a small probability of failing, the chances of the overall project being successful are in fact very poor. Consider, for example, a software program consisting of 1000 lines of code, each of which is 99.9 percent reliable. The chance of the program itself working is only about 36 percent! Brooks' experience has led him to the following rule of thumb for software projects. As a fraction of the total time of the project, planning consumes about T, coding consumes i, component test consumes i, and system test consumes I. Thus, if a project estimate is made based on the expected coding time (the main element for which we can derive an estimate), this in reality will usually represent only about 17 percent of the entire project time rather than the 80 to 90 percent commonly assumed.

The second factor is what Brooks calls "the mythical man month" and relates to our tendency to assume that workers and time are interchangeable. Thus, when a schedule slips, the traditional response is to add labor which is like trying to douse a fire with gasoline. Our assumption that workers and time are interchangeable is correct only when a task can be partitioned such that there is no communication needed between the workers, such as in picking cotton by hand. Most projects, however, especially computer programming, are not set up that way and the more workers that are added require even more workers to train, as well as lines of communi-

cation to coordinate their efforts. Thus, three workers require three times as much pairwise intercommunication as two, and four require six times as much, etc. This result is captured in Brooks' law.- Adding manpower to a late software project makes it later.

Emanon Aircraft is a major manufacturer of aircraft parts, specializing in landing gear parts and assemblies. They are located in a highly industrialized midwestern state. The local area suffers from somewhat higher than average unemployment, partly because Emanon has experienced a downturn in business. In the past three years, they have lost out on a number of landing gear contracts, being underbid by competitors from other areas of the country. Senior management studied the problem, but has come to no conclusion about what can be done. They have hired a consulting team from a nearby university to study the situation and make a recommendation.

Business in the aircraft industry is not significantly different than in many other industries specializing in the building of complex machines. Aircraft builders are primarily assembly operations. They build planes from subassemblies and parts manufactured by themselves or by subcontractors who, in turn, specialize in specific subassemblies; for example, landing gear, avionics, passenger seats, heating and air conditioning, etc. When an order is received to build some number of a given type of plane, the builder (prime contractor) requests bids for the proper number of a certain part or subassembly from appropriate subcontractors. All relevant specifications for the part or subassembly are included in the RFP. The subcontractors who wish to participate in the project submit proposals that include a complete description of the proposed subassem bly together with price information, delivery dates, and any other pertinent conditions of sale.

The university consulting team studied three aspects of Emanon's landing gear operation: the manufacturing process, the cost structure, and the bidding behavior and profit structure on landing gear bids. They determined that the manufacturing process was reasonably efficient and not significantly different from Emanon's competitors. Second, they found that all competitors were using approximately the same level of mark-up when determining their cost-plus price. When examining the cost structure, however, they noted that in the past three years, the firm consistently ran negative cost variances in material accounts. That is, the amount of material actually used in the construction of landing gears was approximately 10 percent less than the plan indicated. The team was unsure of this finding because there were only a few winning contracts for landing gears during the past three years.

An investigation was conducted on the estimation and purchase of materials for this department. It exposed the following facts. Three and one-half years ago, Emanon was late making a delivery of landing gear parts. The firm paid a large penalty and was threatened with loss of further business with the prime contractor. The late delivery resulted when Emanon ordered an insufficient quantity of a special steel alloy used in landing gear struts, and was unable to purchase any on the open market. The steel company re quired a manufacturing lead time of more than 90 days, so Emanon's delivery was late.

As a result, the purchasing official who had responsibility for this contract was demoted. The new purchasing official handled the problem in a straightforward fashion by in flating the material estimates by approximately 10 percent. The cost of material is about half of the total cost of landing gear production, which resulted in bids that were approximately 5 percent above the competition.

Project Management Made Easy

Project Management Made Easy

What you need to know about… Project Management Made Easy! Project management consists of more than just a large building project and can encompass small projects as well. No matter what the size of your project, you need to have some sort of project management. How you manage your project has everything to do with its outcome.

Get My Free Ebook


Post a comment