Estimating

Since I do a fair amount of it, I have been thinking about it quite a bit lately. Its an activity that we learn to do very early, and often do almost subconsciously. It is ingrained as children since it is a natural activity – how much farther is it? When will be there? Children ask for estimates from adults, and they learn to make them fairly early as well. Estimating is a base part of planning, though there is a lot of help out there. For instance, now that internet programs give directions, the estimated amount of time to get from point A to point B is quite easy to determine. But it is the simplicity and “everydayness” about estimating that makes the topic not studied as it should be. We take for granted why we do it and how we do it. I believe that with a little more thought behind what it is, it could make us better about our methodologies.

When we got to the estimating section in class while I was in graduate school there were a couple of notable ideas to take away. First, every person had a different estimate of how long project X should take. Sounds not quite right on one hand, yet, it could be true depending on construction methods and variables determined in the estimating process. Secondly, each persons experience level varies, and the method of estimating varies. These variances create different budgets solely based on previous experience. In this way, the more experienced the estimator is, theoretically the more accurate the estimate is.

Methods of estimating:
Analogy: In the past a flat took 3 hours. This flat is nearly the same it will take about 3 hours also. It cost 200 last time, so we can pencil that amount as well. Or the new flat is slightly more complicated lets multiple labor and materials by 10 percent.

Bottom up: taking an element, engineering it, counting materials, breaking down build tasks and assigning each one an amount of time. (The time inevitable is estimated somehow but that’s a mute point at the moment). Accurate, especially for materials, but takes a long time to finish the estimate. Labor tends, in my experience, to be close as well in terms of the overall project, but less accurate in terms of each task. I think this is mostly due to efficiencies. It may take a half an hour per cut list if you do them individually, but you may be able to do 4 cultists in an hour instead of 2. Secondly, the 1st flat may take 4 hours, but the 10th flat that is the same may take 1 hour. In these circumstances setting up the jig may be as time consuming as welding all of the pieces once the jig is complete. Third, I think we can only estimate so precisely – and that results in a little bit of slop in each number. The more you break it down the more slop happens. For instance think of all the steps to making a pot of coffee. I remember in class we defined some 15 tasks, and it took like 25 minutes for the coffee to be done. Not exactly true to life. While the percentage of inflation that we saw in the pot of coffee example is more exaggerated than in most scenery example, it should be noted. Yet, I have found that it tends to still work out, because the built in contingency helps in the long run for the inevitable snafus that come up.

Units: Similar to analogy in ways, but in this method you count say the number of sheets of luaun needed for all of the flats and you extrapolate information from that base number. Labor might be 2 hours per sheet for easy projects, 4 for more complex shapes, and 6 if there is a laminate or surface that is needed beyond the normal paint treatment. If you know that a 4x8 platform needs uses 36’ feet of 2x4, then 5 platforms will use 180’. This works, except it can get inflated if flats or platforms share members. This estimate is partially based in past experience, but also based on counting. As usual, the hardest part of estimating is the hours. The materials are easier (provided you can count, have some experience in the materials, and understand the level of support needed for the materials).

Square footage pricing: not often done in theatre environments, but practical for early preliminary estimates. I think it would be interesting to take past sets and create square footage costs from them (perhaps divided among major elements like flats and platforms). Then you could use those averages against a new project very early in the project to give the design team an idea of where they are headed in terms of scope. Pricing like this is often used in the construction industry for building (high end museums cost between 300 and 400 dollars a sq ft), and for general estimating (the cost of have a drywall contractor come in put up drywall, tape, etc, costs 8.00 a SF on average).

The above aren’t all the ways, and frankly, when I approach a project I think I tend to come at it from several different angles, which hopefully increases the degree of accuracy in the estimate. But I think that by understanding how you come to an estimate it better prepares you to defend the estimate, as well as improve it.

The other part of estimating is the need of feedback. I think one of the major issues in our industry as Technical Directors is the lack of useful feedback. If you are way over budget, I’m sure that most TD’s hear about it. But they don’t get necessarily enough information to determine the cause. I am a firm believer in the postmortem process after a show. Part of this process is comparing how much was spent in both labor and materials and how that compared to my original estimate. What were the discrepancies? Why did these happen? Where they planned events? How can this information be used in the future? The key here is that the information is only as good as your coding. If nothing is coded, it can only be looked at globally. If your carpenters track material usage and hours on a line item basis you can track it that much closer. Your level of detail will be dependant on a variety of factors, but it should at least be a detail that is considered.

Labor estimating is in particular need for feedback. It’s generally harder to estimate. You can do it on industry averages (great for construction, pretty hard for theatre). Averages could be created for your theatre but that would take development time. Analogies could be made from previous projects. And a lot of times your own experience is used; “I think I could build this wall in a week… yea that seems reasonable.” The flip side of this is that in our industry, labor is the bulk of the budget – especially if it isn’t salaried.

So what’s my point? The next time you put together an estimate think about why you are doing it. Is it just for the boss or are you using it as a tool to help you run your shop? How do you get your numbers? Are there ways to improve their accuracy? Are there methods to track the results of this estimate? Where in the process does it fall – does that affect the level of detail, or the amount of information necessary for an accurate estimate? While an estimate isn’t meant to be the same as a “bid” or an absolute cost for a show, the closer you estimate to the final product the more credible your estimates are. Then when some show comes in that is way to big for your shop, you have credibility stored to be able to ask for additional resources based on your estimate. (Credibility and estimating is another topic I want to explore another day….)


Resources:
http://tinyurl.com/yuo33k
http://www.wbdg.org/design/estimating.php
http://www.directives.doe.gov/pdfs/doe/doetext/neword/430/g4301-1chp15.pdf
http://www.ucalgary.ca/~design/engg251/notes/cost_estimating.pdf