Homer Simpson believes in BIM ROI, do you?

Recently we had a discussion on the Epic BIM LinkedIn group about ROI that made me want to post this. ROI and BIM have been linked for me since the day I left the field and went into the office. It was the first thing I was asked to figure out. "What is the ROI on doing BIM and how do you measure it?" At first I thought this was an easy thing to figure out but since I have not done it yet I was apparently wrong in this assumption. The main issue has been showing the value of problems that you don't actually have. To do this our industry needs to start using different methods of analyzing data other than just observing historical values.

This seems to be a foreign idea in our industry. We typically use empirical methods to collect data sets but we very rarely use any theoretical methods. To understand the difference think about rolling a pair of dice. Empirical probability would analyze previous rolls to help determine the probability of what you will roll in the future. Theoretical probability would analyze all of the possible outcomes of rolling to help determine the probability of what you will roll in the future. Both are equally valid methods.

I think the reason that everyone has had such a hard time with ROI is that the approach is wrong.  We are spending all of our time trying to collect information that is in no way a real indicator of BIM ROI. We are convinced that the secret to BIM's success is locked somewhere in the number of RFI's, the length of the schedule, or in the number of change orders. But no matter what empirical data you bring back it still does not answer the one question the industry really wants to know.

How do you prove that doing the project with BIM brings more value to the project than doing it without BIM?

Everyone knows that there is no empirical data (one-offs don't count) to help answer this question so it is just assumed that there is no possible way to answer the question. But instead we try to force the issue. We analyze one project that has less RFI's than average and proclaim that "BIM reduces RFI's". We do the same thing with Change Orders, Schedule, and anything else we can think of. But the logic is horribly flawed.

You would think that bad ROI's would result in people just disregarding them. But in my last post, I talked about how BIM is often made to sound confusing. So to help simplify things, I found a good example of what almost all examples of BIM ROI sound like to me. Please read the following exchange between Lisa and Homer Simpson:

Lisa: ...I could claim that this rock keeps tigers away.

Homer: Oh, how does it work?

Lisa: It doesn't work.

Homer: Uh-huh.

Lisa: It's just a stupid rock.

Homer: Uh-huh.

Lisa: But I don't see any tigers around, do you?

Homer: Lisa, I want to buy your rock.

In this example BIM is the Rock. The ROI is the fact that you don't see any tigers. Replace the word tigers with RFI's or some other totally inconsequential item and that is how ROI is currently reported for BIM.

Unfortunately this practice regularly leads to one of two results. Either you see the flaw in the logic and don't buy the rock, or you buy the rock and jump into a tiger pit to see if it works. Both of those results are bad for BIM and the industry as a whole.

As usual at Epic BIM, since I brought up the problem I will try to offer a solution. As mentioned above, I don't have one yet, but I feel like the answer lies somewhere in the theoretical approach instead of the empirical approach. Or at the very least, a mix of the two. When we report ROI what we are really trying to do, is give anyone concerned a better feeling out the probable outcome of future work. So just like my dice example, either method can provide a valid answer to the problem. But in the absence of empirical data, we need a more theoretical model. A way of analyzing projects in advance and giving you an idea of the expected outcome using BIM processes and the expected outcome not using BIM processes. This is a very viable option given the correct criteria. The problem is that everyone is so busy collecting useless information that no one (that I know of) is actually looking into what that criteria is.

Just like the dice example, this analysis is more difficult than using empirical data. But like the dice example, the results are just as valid. I know there are many that will say, well who cares we are all going to do BIM anyway because the market place demands it. I am not entirely sure that's true any more. I think lately there have been a lot of people that have become jaded with the BIM. I feel a decided loss of momentum for BIM in the industry. I think that if we don't want the BIM train to come to a halt the experts need to offer a better answer for those that are wondering why we do BIM, other than the rocks they have been selling.


  1. Does the answer lie in tracking? If you track the problems encountered at earlier stages in the project, and assign savings values: i.e. catching the same issue during the construction phase (using historical data) you can accrue an actual value for each project that uses BIM and compare it to similar projects that do not. In addition to design and construction issues, you could also track time savings as mentioned in Antony's post as well.

  2. Thank you for such a great article. The struggle to provide convincing ROI has been the elephant in the room at every client meeting, proposal and academic talk that I’ve been involved with. The fall back is typically anecdotal evidence, which while compelling, is not always convincing.
    The problem is that doing BIM right does and should change the process and this muddies ROI with chicken/egg type discussions. For instance, are RFI/change orders reduced by clash coordination alone, or by the increased communication and collaboration necessary for successful coordination. When done right BIM brings people together. BIM is a catalyst for that value, but not the primary cause.
    Ultimately, my best arguments derive from the intuitive value of BIM. BIM ups everyone’s game. It simply makes problems easier to identify, earlier. Working in 3D shows more and requires more resolution than 2D ever did. Managing model data requires more consistency and detail from the author. Modeling makes teams work together or, when contractually hamstringed, want to work together. These arguments can be hit or miss with numbers focused clients. However, they have a wider target than anecdotes and are more convincing than pulling numbers out of the air.

  3. A theoretical approach gets you a theory. The validity of the theory is proven only by empirical means. That is the trouble with the "soft", or social "sciences". And face it, most of the building construction is a lot more like social science than it is like physics or chemistry. Theories are a dime a dozen in management, economics and sociology, as are truisms and maxims. These rules of thumb are sometimes borne out by experience, but it is very hard to know ahead of time whether conditions this time are such that pulling the same levers one pulled the last time with good results will have the same results this time … assuming the same levers are even there to be pulled. There are just too many significant and uncontrollable variables. Same for construction.
    For this reason, there is no theoretical basis; no foundation on which to build a theoretical argument for (or against) BIM, IPD, CM/'GC, or any other approach to construction or design.
    There is no academic pursuit of a general theory of construction or design. There are only fads. When conditions change, there are new fads.
    The closest we can come to controlling the variables in construction is when the same facility is built repeatedly, and even then, local conditions; the site and all it implies: traffic, neighbors, climate, regulations, real estate markets, supply channels, labor availability and organization all conspire to muck up any message one tries to find by reading the tea leaves of "similar" projects.
    I wish you luck in your pursuit of a theoretical argument for BIM. Maybe it is out there somewhere, but I sure don't know where to look.

    • I was waiting for this argument. I know that this is the way we think in our industry. But you are wrong in your assertion that an empirical approach gets you anything more valid than a theoretical one. Believe it or not, your predictions about the future based on empirical evidence are also just theory. Facts about the past are not facts about the future.

      Take my dice example. Both methods will give you a good guess as to what the future outcome will be. Neither method is more likely to be correct. The only thing that the empirical method could have over the theoretical in that scenario is that if the same set of dice is used the dice may have a bias towards a certain set of numbers. That bias if understood can be built into the theoretical method as well. Hence my suggestion at the end that a combination of both methods may be a good approach.

      I think we can control these variables if we open minds to the idea that this can be done. The issue with ROI for BIM is that I think it has hit an empirical wall and another approach is needed.

      Thank you for wishing me look. Like I mentioned in my article this is a hypothesis and certainly requires some serious testing. Testing I know I don't have the time for.

  4. The other form of analysis is anecdotal. Not robust, I know, but in the absence of blind trials with controls it has some use.

    I know when my team uses BIM I don't have to check drawing cross references, whether sections & elevations match plans, that tags match schedules. And that saves me a ton of very, very tedious work.

    I had a young fellow in a site office get me to talk to his boss about BIM because he was sick of spending literally weeks checking that schedules matched drawings. Something he didn't have to do on his previous job because it was done using BIM software.

    If a contractor asks for a section through the building I can provide it in 10 minutes, not a day later.

    If the client wants an updated area schedule I just print it and send it, I don't have to get someone to manually measure every space and add them up.

    All these things MUST save money because they save time. Well, in theory, as an architect, all it means is I spend more time designing. But at least I enjoy that.

Leave a reply