An engineer at a very large corporation recently sent me an email in response to my estimating software post. He felt that I did not address the up-front estimating requirements of building owners prior to formal design. That is, how do they create a conceptual estimate prior to having detailed drawings and bids? Moreover, how do they compare these initial estimates to the actual costs when the project is complete?

The engineer is looking for a better way to facilitate up-front estimates during the screening phase based on utility system costs and assembly type estimates before detailed drawings are available. His major challenge is balancing the need for speed with the need to minimize risk. Currently, he believes that his accuracy is +/- 30%.

Later, when the project has been approved for development, his firm moves into the concept development stage and seeks to create a more detailed estimate for funding approval.  This involves a fair amount of pre-engineering work, but this is still not at a working drawing stage. At this stage, the estimates have an accuracy of +/- 10% - 15%.

Only after that stage does the firm move into the final design stage that creates the detailed drawings from which they can let the project for bid. At that point, they can use the traditional takeoff and cost estimating systems I discussed in my Preconstruction Management Software Trends and Strategy report and use a contingency of 5%-10%.

But how can he improve on his early estimates? At present, he develops those estimates and then has no database to capture the information. Instead, he creates Excel spreadsheets that are later placed in binders or stored electronically in Excel. When a new project comes along they are not able to easily use data from the old projects because it is not in a well-structured database. The challenge is complicated by international locations, currency fluctuations, inflation and raw materials prices. Finally, the cost data that comes after project completion tends to be presented in a line item CSI type format (e.g., all drywall used) consolidating all similar items rather than spread out in the assembly format used in the initial estimates.

I would be curious to hear readers' responses to this challenge. What techniques and systems have you seen that address this challenge? Are there any software vendors that have done a particularly good job addressing this initial estimate process? Has anyone built their own system in house?

March 12 , 2007

Model based conceptual estimating is the way to go. I've tried Revit and Graphisoft, I like Graphisoft better because it is true 5-D, cost and schedule.

David Wilkinson

February 16, 2007

You have touched on an age old problem. I have been estimating for over 20 years myself and the issue never changes. The issue is not a matter of software. That is kind of like asking a carpenter which hammer handle (fiberglass, wood or steel)builds the best house. It is not the hammer, it is the user of the hammer and his skill and attention to detail. At our firm we have done a pretty good job of tracking estimated versus actual costs for our projects. Fortunately for us mmost of our projects are within a 200 mile radius and many are similar types of buildings. We were experiencing a pretty good record of hitting our initial forcasts (usually within 10% of Schematic Estimates and within 5% of Design Development estimates). However, the recent spikes in materials and some more challenging projects put that record at risk. However, we learned from our experiences and moved forward and are now back on track. The answer to your questions is first and foremost, "Experience". Nothing replaces "Been There, Done That". Track your accuracy across a variety of projects. When you have a bust dig deeper to try to understand why. And share your experiences with others. Also don't be afraid to rely on others with specialized knowledge when you are in a realm that you have limited or no experience. The bottom line is that you have to always remain vigilant and on your toes. This business has a way of keeping us all very humble whenever we become too confident or too relaxed.

Walt Tack
Reynolds Construction

February 16, 2007

The best answer is "history" (from actual projects or definitive estimates). But no two projects are ever alike. And even if they were, who in my business has the spare time and budget to organize, store, retrieve and analyze that information? Unfortunately, in the real construction world, it's always "on to the next project".

Have you ever actually done construction estimating as a construction professional? I've been doing front-line construction estimating, including D/B, for over thirty-five years and have the white hair to prove it.

If you ever find that magic pill you're looking for, please let us all know.

Frank Sifford
Manager Construction Services
Black & Veatch Construction, Inc.
Charlotte, NC

February 14, 2007

I came across your article by accident. I don't want to sound like a commercial but Eos Group Inc. has a product called Eos Advisor that address this problem. Advisor allows companies to compile similar data from both estimates and as built data. The data can be adjusted for differences in timing of projects and for locations. For more information our website is www.eosgroup.com.

Ann Kenkel
Eos Group Inc.

February 14, 2007

Not to beat a dead horse… but Excel is known as a "spreadsheet/calculation" application whereas Excels close cousin Access is the "database" workhorse (that also handles calculations nicely). It surprises me how so many people are comfrotable with Excel while leaving Access sitting on the shelf as it were. Give me anyone "storing" data in Excel and in minutes I can provide a greatly improved version in Access.(and I am in no way an expert!) Of course, any Excel database can be imported into Access.

Ken

February 13, 2007

The most promising program I have seen is the new modeling, 5-D, programs that enable the user to build conceptual estimates from their existing database of material and labor pricing. The problem is that we do not do a good job of building effective DB to support the significant capabilities of the modeling software. In Europe the contractors do a better job of building and maintaining productivity information. The material cost is relatively easy bit labor/productivity is very difficult because we tend to think 'It is what it is' and do not try and capture pre/post data for use in conceptual estimating. The new modeling programs may drive us to that, I hope so!

David Wilkinson