Marketing Will Deliver Damaged Goods

By ROBERT ZIMMERMAN

Cigarette companies used to tell us that 9 out of 10 doctors smoke Chesterfields. And who could forget “the world’s most interesting man” choosing Dos Equis?

These bold statements were proliferated long before the inundation of absolute trash we now scroll past on social media every day.

It comes down to our insatiable need for attention. We will say just about anything if it benefits our agenda.

So when a construction tech cheerleader tells you his or her AI tool will save you 30 percent to 40 percent on labor costs, cut your project timeline in half and practically run your back office for you, I want you to hear that the same way you’d hear a cigarette company telling you its product comes “doctor-recommended.”

Because right now, in construction, AI is the Chesterfield cigarette.

Everyone’s selling it. Everyone’s claiming it. And almost nobody is asking the question that actually matters: will it kill my business if it’s wrong?

You Can’t Build on a Broken Foundation

I’ve been having real conversations lately with real operators. Not pitch decks. Not conference panels.

Actual conversations with people who are in the middle of this every day.

Landon Messal of SiteMarker laid it out plainly: construction projects are starting before the plans are finished. Yellow iron is hitting the field while the design team is still figuring out what the final design looks like. We’re pricing incomplete drawings just to get into the municipal review queue. We’re moving dirt while the permits are still in process.

This isn’t new. But here’s what is new: we’re now trying to plug AI into that fairytale of a project plan or scope of work and expecting it to produce the next golden calf of construction lore.

Bad Idea.

Every scheduling tool, every cost tracking system, every labor forecasting platform you own is only as good as what goes into it. When your data starts on an assumption instead of a confirmed plan, you haven’t built an AI-powered construction company. You’ve built a faster way to be confidently wrong.

Messal made another point that stuck with me: construction industry productivity has grown at 0.4 percent compounded annually for two decades. Every other sector combined has grown five times faster. We didn’t lose that ground because we lacked technology. We lost it because we never fixed the underlying processes that technology was supposed to improve.

AI doesn’t fix broken processes. It inherits them.

The Most Expensive Spreadsheet in Your Business Is Your Payroll

I also sat down with Anna Berger, co-founder of Trayd, and she said something I haven’t been able to stop thinking about.

Labor is the single-largest cost center for most specialty contractors. And the majority of those contractors are still managing it with paper timecards and Excel spreadsheets. Payroll admins are spending 12 to 14 hours processing a single pay cycle, before a single number hits the payroll table. Then they wake up on Monday and do it all over again. Seven-day cycle. No room to breathe. No time to fix what’s broken. Just survive the week and start over.

One bad paycheck costs upwards of $960 to correct. Compliance errors on certified payroll go directly to the Dept. of Labor. The fines make $960 look like a rounding error.

Welcome to unstructured data. Paper that someone has to read, interpret and manually key into a system. Spreadsheets that live on someone’s desktop and die when they quit. Disconnected tools that don’t talk to each other and never will because data governance sounds more like legal jargon than an accountability tool.

Berger said it directly: until you structure the unstructured data, there’s not much AI can actually do for this industry. You must clean up the house before you can leverage the tools everyone’s promising will transform your business.

She’s right. And most of the industry isn’t ready to hear it.

The Pattern Nobody Wants to Admit

500 companies. 1,000 completely different problems. One root cause.

Data that was never properly captured. Processes never designed to feed clean information downstream. Systems built for every industry but ours, jammed into workflows they were never meant to support.

I’ll say it plainly: AI is not a strategy. It is a drunken Einstein. Feed it clean, structured, trustworthy data and it will accelerate everything – planning, estimating, labor forecasting, compliance, productivity. Feed it paper timecards, overlapping phases and 14 spreadsheets that don’t agree with each other, and it will give you faster, more confident, more expensive wrong answers.

I know this is a gut check moment – but this is where the biggest gains begin.

Where This Actually Starts

You don’t need a transformation initiative. Those are overrated and cause panics among the troops. But you do need to ask one honest question inside your own organization:

Where does our data actually come from,  and do we trust it?

Start in the field. How is time being captured? Is someone writing it on paper at the end of a shift, handing it to a foreman, who hands it to an admin, who keys it into Excel? That’s not a data system. That’s a game of telephone with weak Vegas odds.

Then work backwards. Where does information break down between the field and the office? Where does someone have to make a judgment call because the source data isn’t reliable? Where are your systems advising decisions based on inputs nobody has actually verified? I would definitely be real critical of “The Schedule.”

That’s your data governance problem.

And it was there long before AI showed up to make it someone else’s selling point.

The construction companies that win the next decade won’t be the ones who bought AI first. They’ll be the ones who built the foundation that makes AI actually work.

That’s not a marketing line.

That’s just the truth. And unlike the cigarette companies, I don’t need a doctor to back me up.

Robert Zimmerman is host of the ConTech Exec Podcast and an executive consultant to C-suite leaders in construction and construction technology.