Types of Estimates, Tools, Data Sources, Methods

A sound understanding of the common types of estimates, tools for estimating, data sources, and methods of estimating forms the basis for more sophisticated estimating concepts. To the novice, this chapter presents the fundamentals of estimating as it is practiced in construction today. For the construction professional, the chapter serves as a review of effective methods. The remainder of Part I is an expansion of the introductory information presented here.

Estimators typically perform at least five different types of estimates, depending on their employer and position. These include preliminary or ballpark estimates, square foot estimates, assembly or conceptual estimates, contractor's bid estimates, and engineer's or owner's estimates. Although these types are generally known by these terms, there are not always clearly defined boundaries between each type. Figure 1.1 shows the general progression from the least to the most accurate estimating types.

Cost analyses for alternative construction methods and materials, and change order and claim estimates are two special types of estimates not shown in Figure 1.1. Life cycle cost analyses, value engineering, and uncertainty analyses are additional types that do not fit the generalized picture. Each of these will be treated in later chapters.

Early in the planning stages of a project, both owners and designers need, and often demand, an indication of the cost of the project at completion. Since the project has not yet been designed, estimates prepared at this time have a low expectation of accuracy. Names sometimes used to describe such estimates include preliminary, ballpark, blue sky, seat-of-the-pants, and order of magnitude. One of the important purposes of these preliminary estimates is to provide information to an owner or client so that a decision can be made to proceed with the project. Preliminary estimates are often made in less than two hours. The expected accuracy is no better than within 20-30%.

Project comparison is a method of preparing preliminary estimates when cost data for comparable completed projects is available. If the comparison project differs substantially in size, adjustments need to be made for the economy of scale. Chapter 2 addresses preliminary estimates in detail.

Square foot estimating is another method of developing a preliminary estimate based on historical data. This technique is effective in preparing advanced preliminary estimates when features of the proposed project are known, but not yet designed. To use this method, it is only necessary to know the type of building or facility and the proposed number of square feet. Other unit measures such as the number of beds for a hospital, or number of spaces for a parking garage are sometimes used instead of square feet, but the basic method for determining these estimates is the same. Annual cost data books are published that provide square foot and other unit costs for a variety of buildings. Square foot estimates are further discussed in Chapter 3.

|Assemblies| are groups of the work of several trades combined into a single cost element. For example, construction of a bathroom may require work by plumbers, electricians, tile setters, and carpenters. The assembly cost includes the cost of participation from each trade, including materials. Publishers of cost data produce books and computerized data tabulating assemblies costs that can be used in intermediate or conceptual cost estimates. |Systems| is another term sometimes used for the grouping of cost data into a single estimating unit. See Chapter 4 for a more detailed discussion of the assemblies method.

Assemblies costing is also used in detailed unit price estimates where they may be referred to as Work packages|work packages.| Here, each contributing trade cost is fine-tuned to the local area and type of project. This is a powerful tool that both saves time and improves accuracy for companies doing repetitive work on the same types of projects. A detailed discussion of the preparation and use of this type of assembly for unit price estimating is given in Chapter 7.

Final detailed estimates can be prepared when the design and contract documents have been completed, or are essentially complete. Generally two types of final estimates are prepared; one by the contractor intending to bid the project, and the other for the owner's use in preparing budgets and evaluating bids for the project. The estimate prepared for the owner is called the |Engineer's Estimate| when the owner is a public agency, and is almost always required for bid comparisons. The owner's estimate is optional for privately funded projects. The major difference between a contractor's bid estimate and an engineer's estimate lies in the information the two estimators have at their disposal. As reviewed in the Introduction (Figure I.1) the Estimatorcontactor\scontractor's estimator normally has much more detailed information at his disposal. The engineer's estimator is forced to rely on more general information, since he does not know which contractor will receive the award and construct the project.

In either case, the final estimate must be as accurate as possible, since it represents the money that will be received or spent when the project is actually constructed. Three methods are often used for these final estimates: unit price estimates, schedule estimates, and previous bid tabulations. These will be discussed later in this chapter.

Contractor's estimators use a variety of methods to produce their bids: some use spreadsheets, others use specialized estimating software. Estimators preparing bids for government or municipal authority heavy construction projects may combine the proposed project schedule and the estimate, since meeting the scheduled completion date is often required by the contract.

Regardless of the method the estimator chooses, they all have one thing in common: Utmost attention must be paid to every detail, and multiple checks are necessary to ensure that no item of work is omitted or duplicated. (Chapter 6 covers this type estimate in more detail.)

Engineer's or owner's estimates are used both as a basis to evaluate contractor's bids and to provide the owner with the budget needed to pay for the construction project. Most construction administrators believe that the engineer's estimate should not match the low bid, but should fall somewhere between the second and third bidders. An analysis of historical data bears this out. Approximately 235 bid openings were analyzed where both the engineer's estimate and the low bid were known (based on bid tabulation data published in Engineering News Record over a period of about eight years). Figure 1.2 is a histogram showing the distribution of the percent deviation of the low bid from the engineer's estimate. A statistical analysis of the data suggests that the average low bid is 7.6% below the engineer's estimate. These results agree with a comprehensive analysis of engineer's estimates in Europe, Canada, and the United States prepared by Dr. Martin Skitmore of the University of Salford, UK, |Factors Affecting Accuracy of Engineers' Estimates.

Dr. Skitmore's analysis shows that low bids generally fall between 5% and 15% below engineers' estimates. Individual estimators' accuracies range from 8% to 16%, with standard deviations ranging from 9% to 31%. These results support the belief that engineer's estimates are simply not as accurate as those prepared by contractors for bidding purposes. As shown in the introduction, the engineer's, or owner's, estimator is not privy to the detailed information held by the contractor.

There is usually more than one way to accomplish any given construction task. Alternatives of construction methods, crew, equipment, and material often need to be evaluated to determine which is least expensive, or best matches the project or company objectives. This process should be considered an exciting opportunity rather than a burden. Industry progress can only occur through the adoption of more efficient or more effective methods and materials. In order to test an innovative method on an actual project, however, detailed cost and performance predictions are almost always required to persuade management.

Changes in construction projects are inevitable. No matter how well planned a project may be, there are always a certain number of changes. Each change, whether proposed by the owner or the contractor, requires documentation, negotiation, and an agreement on cost and schedule. The contractor's estimator may provide the cost and schedule projections for the change order proposal, or, as is often the case, the project manager is charged with this task. Here's where the architect or engineer's estimator steps in again to analyze the contractor's change order proposal to decide whether it is "fair and reasonable" within the context of the contract. Chapter 16 provides additional detail on change order and claims estimating.

The rapid development of microcomputers, associated cost estimating and project management software, and accessories such as digitizers and plotters, has made a major impact on construction estimating. Estimating technology has advanced significantly in four areas: quantity survey or Estimatingtoolstakeoff, estimating software, computerized data bases, and availability of unit cost data on computer media. These time-saving advances, described briefly in the following paragraphs, enable contractors to bid on many more jobs than were previously possible with manual estimating. More detail on automated estimating is presented in Chapter 7.

The development of computer aided drafting (CAD) has revolutionized the design/drafting industry. Data previously available only on paper (the drawings) is now duplicated in digital format on magnetic media. Electronic digitizers enable transfer of measurements directly from the plans into the computer, where programs convert dimensions to the quantities required for estimating. Some cost estimating software products accept data directly from drafting software (CAD) products. This enables direct transfer of dimension and quantity data from the project plans Quantity surveyto the computerized estimating system. As a result, a takeoff formerly requiring weeks can now be done in hours. Future advances will likely link such data directly to project management and other software systems.

Even the simple quantity survey tools have advanced. The time proven polar planimeter used to measure areas on drawings is now available with built-in scale factors that will present area measurements in the units needed, such as square feet, square yards, or square meters instead of square inches or square centimeters. A few of the more advanced planimeters can transmit area measurements directly into a computer for use in an estimating program. Modern versions of the simple roller wheel for measuring length of lines are now battery powered and can be rolled along the perimeter of the plans to read lengths in feet, yards, or meters, once the drawing scale is entered into the instrument.

Spreadsheets have been used for estimating for decades. Most cost data publishers, as well as architectural, engineering, and contracting companies use printed estimating spreadsheet forms, although they usually go by another name ("estimate sheet," "estimate summary," "Materials listmaterials list," etc.). Although these are, in some ways, the spreadsheet of the past, they still serve a useful purpose. Some calculations and estimates are simply too small to bother with a computer. The forms also serve as handy worksheets for recording data gathered out of the office or over the telephone.

For larger projects, currently available spreadsheet software presents the well known columns and rows with automated calculations and a host of additional capabilities. This now makes alternative analyses (the "what if?" feature) possible. Analyzing the cost ramifications of many different methods and materials are too time consuming to consider without a computerized spreadsheet. Many design and construction professionals today calculate all their estimates on personal computer-based spreadsheets. Spreadsheets that import cost data from a base of construction costs are fast becoming the trend of the future.

Well-designed spreadsheets organize the estimate and present it in a neat and orderly manner, whether done with a pencil on a data pad or with a computer. The advantage of using the computerized spreadsheet is that it virtually eliminates computation errors.

Additional applications of spreadsheets may well be more important than the preparation of complete estimates. Spreadsheets can be created for a myriad of supporting calculations such that they are a ready time saver. This alone makes the spreadsheet the "estimator's workhorse." Chapter 7 shows several typical applications.

A feature article published in ENR (formerly Engineering News Record) in September 1989 stated that 173 software packages for computerized estimating were available2. This number is growing constantly.

The basic components of a complete estimating software package are a database of costs coupled with a calculation module. The simplest programs contain little more. The power of the system increases when it is interfaced with other software for importing quantity data directly from drafting (CAD) packages or digitizers, or for exporting estimate data to accounting, scheduling, or other project management packages. Cost and complexity Estimating software also increase when these interfacing features are included. The time required to become proficient with the software also generally increases with the complexity.

At least two Federal agencies, the Naval Facilities Engineering Command (NAVFAC) and the U. S. Corps of Engineers, require both designers and bidding contractors to prepare and submit estimates on their specified system. Both of these estimating software systems, complete with databases, are available on optical disks (CD-ROM) from the National Institute of Building Sciences in Washington, DC4. Both the Corps of Engineers and NAVFAC design manuals and specifications are also included as part of the set.

Project management programs were originally limited to scheduling, usually either Critical Path Method (CPM) or Program Evaluation and Review Technique (PERT). Current programs can account for resources, both fixed and variable, and provide many more features. Project management software today implements a complete management system, including scheduling, cost estimating, cash flow projection, and tracking.

Database software is much more efficient than spreadsheets in managing the large volumes of cost data needed for estimating. Database software packages have evolved to the point where both the price database and the calculations required for estimating can be included in the same package. Recent developments in both spreadsheet and database software enable data transfer between the two programs. This allows each software program to perform the function it does best.

All estimating methods, whether manual or computerized, require cost data. The data may include labor wage rates, equipment rates, material prices, complete compilations of unit prices, time required (duration) to complete previous projects or tasks, tabulations of bids submitted for previous projects, and markup or overhead information. All of the data needed for small projects might be kept in a looseleaf notebook. However, most estimators dealing with medium- to large-size projects need databases consisting of thousands of items. The various types of data sources are described in the following paragraphs.

Most suppliers of estimating software supply a cost database, normally of unit prices, for initial use with the product. The database may contain national averages, or may be based on the developer's city or region. In either case, it is necessary to modify the labor, equipment, and materials data to fit the estimator's own locale. Implementation of this type of system can be a major undertaking; to the point where packages are occasionally purchased and then never used properly.

Estimators and managers of estimating departments must realize that the cost of maintaining current valid cost data is the largest cost element in procuring estimating software. The cost of maintaining an accurate current cost database may actually exceed the cost of retaining a senior engineer or cost estimator on the payroll to provide these functions instead. The advantages and disadvantages of purchasing estimating software Data sourcesshould be carefully weighed...what are you really getting for the cost? Roughly eight to ten companies compile, publish, and market books containing cost data. Types of data include material, equipment, labor, individual line item unit costs, assemblies unit costs, and others. At least two of these companies now prepare and market specialized databases to vendors of estimating software. RSMeans Co. offers inexpensive versions of their annual cost data books on magnetic media (disks) for computerized applications.

Unit price Bid openings from previous bid openings can be obtained upon request for public works and heavy construction projects, although generally not for private projects. Previous bid tabulations can be especially useful to engineer's estimators, and are readily available since owners, especially those in the public domain, retain an archive of past bid data. Generally the data from previous bids should be limited to the lowest three bidders for any given project. The higher bidders may not have understood the job requirements or they might have been merely submitting courtesy bids.

Local bid tabulation data from state, county, and city jurisdictions are usually available to any estimator who makes a serious effort to obtain them. More detail on application of previous bid tabulations is provided in Chapter 14.

Estimating methods and procedures vary according to the type of estimate prepared and the quantity and quality of design information available to the estimator. These range from very simple procedures which can be accomplished in a hour or so for preliminary estimates, to complex measurements and analyses requiring weeks for complex projects.

Initial preliminary estimates are often requested and prepared with no design information other than a general concept, such as a "three-story building of about 60,000 square feet." Obviously, for such a rough description, only a "ballpark" or order of magnitude estimate can be prepared. This can be done using a project comparison, as explained in Chapter 2, or square foot type estimate as explained in Chapter 3.

Intermediate or conceptual estimates require additional information about the project, although very little actual design work may have been done. At this stage, the proposed three-story building may be further described. For example, it might have:

• 20,000 square foot area for each floor
• Basement parking equal to the floor area
• One elevator
• Concrete column construction with elevated floor slabs
• Curtain wall
• Drywall interior

This higher degree of detail allows estimators with the above basic information to use published assemblies and systems cost data. Like the unit price data tabulations, these prices can then be factored to account for differences between the base used in the published data book and local area costs. Assembly estimating is explained in Chapter 4.

Complete or nearly complete contract documents are necessary to produce the final detailed estimate. Plans and technical specifications must be available to the estimator. The general estimating procedure for unit price estimating is:

• Prepare or obtain a quantity survey.
• Divide the work into individual line items or assemblies according to the coding system used (typically MASTERFORMAT).
• Obtain unit prices for each line item or assembly and calculate the extension to obtain the line item or assembly costs.
• Decide or obtain all markups, overhead, and profit factors.
• Group into bid or invoice items according to the type of contract.
• Arrange the estimate into the format required for bid or submittal.

This general procedure is modified when a combined schedule and estimate is prepared using project management software. Resource groupings (crews), task durations, and material costs replace the line items or assemblies shown above. Additional manipulation may be required to handle markup, overhead, and profit depending on the software available and the type of bid or submittal required. Chapter 6 provides a detailed overview of both unit price and schedule estimates.

Pareto's Law (Vilfredo Pareto, Italian economist 1848-1923) states that 80% of the effort is spent on 20% of the work. In cost estimating, this is sometimes known as the "Law of Triviality" in which 80% of the effort is spent on the 20% of the items contributing the lowest cost. Likewise, when an estimate is nearly 90% complete, the estimator may then spend much too much time perfecting the final 10%. This is the stage at which estimate preparation runs over the budgeted time period, and estimators should be aware of this trap. When starting an estimate, the estimator should search for the few items with the highest cost, and concentrate on them. The remaining time is devoted to items of decreasing priority. Many estimates have a few insignificant items which contribute only a fraction of 1% to the total project cost. A standard figure, or even a wild guess, is adequate for these very low cost items. It is much better to spend the time where the dollars are.

This chapter presented types of estimates, tools, data sources, and methods of preparing estimates, and showed in general terms when each type should be used. There is considerable overlap, both in methods and preparation time, for both preliminary and intermediate estimates. Likewise, a variety of tools, data sources, and methods are available for the estimator's consideration. The choices for any one estimate depend on the project, degree of accuracy required, time given the estimator for preparation, and the data sources and tools available.

Following chapters (Chapters 2 through 4 and Chapter 6) show the estimating methods in increasing detail. Chapter 5 provides a detailed discussion of data sources and methods.

Chart a sequence of cost estimates needed from the time a client wants "a three-story office building" to the contractor who bids on the project.

Given an engineer's estimate of $10,000,000, what would a typical first (low), second, and third bid be?

When would a schedule estimate prepared with project management software be preferred over a unit price estimate?

What are the advantages and disadvantages of using published cost data, either on paper or on disk? (Save your answer, and compare it to the same question in Chapter 5.)

Martin Skitmore, Factors Affecting Accuracy of Engineer's Estimates, Cost Engineering, v30, n10, October, 1988, pp 16-23, American Association of Cost Engineers, Morgantown, West Virginia.

Estimators Evaluate Automation, pp. 32-34, Engineering News-Record, September 14, 1989, McGraw-Hill, New York.

The Bid Reporter, Semi-Monthly, (Alabama, Florida, Georgia, Illinois, Louisiana, Mississippi, North Carolina, South Carolina, Tennessee, Virginia), The Bid Reporter, 1109 Alpine Hills Court, Stone Mountain, GA 30083.

Construction Criteria Base, Annual Subscription, National Institute of Building Sciences, 1201 L Street, N.W., Suite 400, Washington, DC 20005. 1.1 Relative Accuracy of Estimate Types

1.2 Percent Variation from Low Bid