Cost Estimating: Art or Science?
A few years ago, I attended a big construction conference in San Francisco with some of my coworkers in the estimating department. I won’t forget one of the presentations about the future of technology and AI in construction. One of the slides had a list of job functions in the industry that are expected to be completely replaced by AI at some point in the future. Guess what was on the list: cost estimator! So, if you’re an estimator, look for a different job because your job won’t exist in the future.
I strongly disagree with that, not just because I worked in an estimating department at the time. There are many such predictions out there that are based on poor judgment and lack deep domain knowledge. Have you noticed they all predict something crazy happening at only perfectly round number (2040, 2050,…) years? Today, I run a cost estimating technology startup and still believe that cost estimators are not going to be replaced by AI. Technology will not replace estimators but will empower them. It will automate the repetitive part of their work and enables them to spend more of their time analyzing the estimate and refining it based on their experience rather than performing tedious and repetitive tasks.
In order to develop an accurate estimate, you need both science and art of estimating. We, at Zebel, have developed a technology that automates the science part and allows the estimator to add the art part and develop an accurate estimate in minutes. Our estimating technology has been tested and validated in many projects of large and well-known developers.
Cost estimating is an important part of pre-construction work for any project. It may seem like a repetitive work that involves not so complex calculations that boils down to measuring quantities of different elements of the projects (concrete volume, steel weight, windows and doors count, etc.) and applying current market pricing to those quantities. That’s a naive view to cost estimating that only looks at the math part and ignores the “art” side of it. However, if you talk to any estimator, they’ll tell you how important the art part is. Measuring quantities may seem like simple task but it takes a lot of experience to know what to measure and how to do it. Estimators have to make many assumptions to develop a cost estimate and they draw from their experience to make reasonable assumptions. Too often, current market pricing may not be readily available and you have to use historical pricing.
Imagine you want to estimate the cost of cabinets for your new 100 unit, 85,000 square-foot apartment building project. You have schematic drawings for your project that do not really include much details about the cabinet design and specifications. So, no cabinet sub-contractor can give you a hard bid. You have no choice but to lookup your historical cost. Your cost of cabinets for your last project (150 units, 140,000 square-foot) was $600,000. How do you use that piece of information though? You may calculate the cost per apartment unit of $4,000 ($600,000 / 150) or cost per square foot of $4.28 ($600,000 / 140,000) and apply that to the new project. But depending on which unit of measure you choose, your estimate would be either $400,000 ($4,000 x 100) or $363,800 ($4.28 x 85,000). The two estimates are 10% apart from each other. Which one is accurate? The correct answer is none of them. There are other factors that should be taken into account as well.
Neither unit count nor square footage are appropriate unit of measures to estimate the cost of cabinets. The two projects have very different unit mixes. The historical project had a lot more studios (70 studios, 60 one bedrooms, and 20 two bedrooms) and while the new project has many 2 bedrooms and no studios at all (0 studios, 40 one bedrooms, 60 two bedrooms). So, there is more linear footage of cabinets in each unit on average in the new project, which makes the cost of cabinets higher than the historical project. Linear footage of the cabinet is a much better unit of measure than unit count or square footage that gives you a much more accurate estimate. But where do you get the linear footage values? You may assume a linear footage of cabinet for each unit type (e.g. 20 ft for studios, 28 ft for 1 bedrooms, and 35 ft for 2 bedrooms) and estimate the total linear footage of cabinet in the entire project (0 x 20 + 40 x 28 + 60 x 35 = 3,220 linear feet) and do the same for the historical project (70 x 20 + 60 x 28 + 20 x 35 =3,780 linear feet) to be able to make an apples-to-apples comparison. Historical cost of cabinet will be calculated as $158 ($600,000 / 3,780) per linear foot. Applying it to the new project gives us an estimate of $511,111 ($158 x 3,220). Now you can see how off both earlier estimates ($400,000 and $363,000) were, when you ignored the unit mix differences.
Also, the historical project was a luxury apartment building with high end finishes, including the cabinets but the new one is more of a work force housing project with average level of specifications. So, the material used in the old project was more expensive. You may apply a 15% discount for the lower end finish level and adjust the estimate from $511,111 to $434,444.
Moreover, the new project has a new component that impacts the cost of cabinets but didn’t exist in the historical project. There is a movable kitchen island in all the two bedrooms that your company never had in your past projects. You may carry an allowance of $1,000 per kitchen island to account for the extra cost of cabinets that didn’t exist in the historical pricing. Adding $60,000 (60 two bedrooms x $1,000) would increase the estimate to $494,444.
Above all of that we forgot a key factor: time. The old project was built 2 years ago and prices have gone up since then. Assuming a 5% cost escalation per year, you may apply a 10% premium to the historical pricing. After adding the cost escalation, your cost estimate is $543,888.
As you can see,the arithmetic calculations involved in each step were simple science but there are many of them and it takes time. This example was just one out of hundreds of items of a typical construction budget. Some of the items are even more complicated and involve more factors to account for. Manually going through that process for the entire budget is very time consuming and that’s part of the reason why contractors ask for a long time to deliver a cost estimate. The knowledge behind performing each step is the art part of estimating that requires experience.
Zebel technology includes both science and art of estimating. We have developed the technology that is analytically sophisticated, does all the data gathering and calculations while it provides a simple interface for estimators to add their input and create a refined budget in minutes. Check out our solutions at www.zebel.io.
