On the surface, construction looks like a race to the lowest bidder. In reality, it is a “trust but verify” economy. Reputable firms rarely compete solely on price; they charge a premium for certainty. Preferred vendor or not, managing risk in construction creates this large layer of “shadow work” that’s been expanding. No wonder we haven’t seen productivity gains. I uniformly hear that drawing quality is going down, in-office work going up, yet the field is frustrated with the in-office outcomes.
Take estimating as an example. You’d think one price estimate should suffice, but that’s never the case. Owners pay GCs to manage costs, while also hiring in-house estimators to check the checkers. It seems inefficient until you look at the ROI: one owner I spoke with saved $40k simply by spending four hours verifying a bid. We have the owner, GC, subcontractor and vendors all do their own estimates. Another interesting trend I’ve observed is that in some trades, the estimating liability has been pushed to the vendor.
A similar dynamic is playing out in design. As the fidelity of original design drawings fluctuates, subcontractors are quietly transforming into engineering firms. This isn’t just about value engineering. It is defensive. In vertical construction, subs routinely take Level 300 BIM models and upgrade them to Level 400 to make them buildable. In heavy civil, teams manually overlay GPS data onto 2D drawings because the field equipment runs on digital models the design team didn’t provide. With quality of design deteriorating, these in-house design workflows turn from getting things field ready to re-creating error free designs. If they don’t fix gross errors, the field crew builds the wrong thing.
Many firms are attempting to solve this by incubating their own software. All seek efficiency gains, with some aspiring to commercialize the software. They view their craftsmanship and their processes as intellectual property, so commercializing is a way to diversify and capture value on that knowledge. This makes sense, except commercializing software is fundamentally a different type of business. While construction is geography, equipment and labor constrained, software isn’t. It’s all about building and deploying to a wide audience, because you can with a press of a button. This presents an interesting challenge for me as somebody trying to build something new for the industry. Each firm wants the benefit - but doesn’t want their rival to have it. Thankfully, there are people that see this as the future and are happy to help get us there.
This work is exacerbated by the constant flux in design. Having built software for so long, I view design drawings as code. I can’t fathom deploying software that doesn’t compile or lacks unit tests. Yet, we erect the world’s most complex structures using designs that require manual verification to function. Imagine a world where every design was instantly verified against building codes, priced, and ready for the field. In that world, I believe we would make a dent in construction productivity.