Your Best Estimator Is About to Retire. Now What?

There are 221,400 cost estimators working in the US right now. Over a fifth of the construction workforce is 55 or older. The BLS projects 16,900 estimator openings every year through 2034, almost entirely from retirements and people leaving the field. And the pipeline to replace them is slow. ASPE requires five years of experience before you can even sit for professional certification. Most people in the industry will tell you it takes closer to eight or ten years before an estimator is truly running on their own.

I've spent the last few months calling GCs, architects, and subs, asking questions about how preconstruction actually works. And one theme keeps showing up that I think the industry is underestimating: the amount of critical knowledge that lives exclusively inside experienced estimators' heads.

Not in a system. Not in a database. In their heads.

What "Institutional Knowledge" Actually Looks Like

I'll give you a concrete example. I was on a call with a senior estimator at a Top 10 Southwest GC. After we talked through his workflow, he sent me a Google Doc. It was a detailed breakdown of the most common scope gaps in commercial construction, places where work falls between trades and gets missed.

Power to owner equipment. Interior utility trenching. Gypcrete under tub/shower units. Coring for MEP trades. HVAC commissioning. Condensate drainage. Non-structural blocking. BMS vs. equipment controls. Flashings. Firestopping. The list went on.

Each entry described why it happens, who it usually gets assigned to, and how to prevent it. This document represented decades of experience distilled into a few pages. The kind of thing you only know because you've seen a project go sideways when nobody picked up the coring scope, or when the BACnet gateway fell between the mechanical sub and the controls contractor.

Most estimators carry this knowledge around and never write it down. They just know. They know which subs in their market tend to miss certain line items. They know how their company splits concrete between structural, mechanical pads, and flatwork. They know that when an electrical sub comes in 20% below everyone else, something is probably missing from their bid.

When that person retires, all of it walks out the door.

The Busy Work Is Burying Them

Here's what makes this worse. The experienced estimators who carry all this knowledge? They're spending most of their time on work that doesn't require any of it.

I shadowed a preconstruction team at a mid-size GC and tracked where their hours actually go on a typical project. Scope sheet creation ate 37.5 hours. Subcontractor outreach and follow-up took 120. Bid receipt and processing, another 25. Populating the estimate sheet, 8. Leveling meetings, 6. Add it up and you're looking at roughly 197 hours per project. This team bids close to 100 projects a year.

The guy running this process put it simply: "You take a call while you're driving. You get a call from another trade and your mental capacity totally shifts. You're going from electrical to plumbing. Then you're closing out another job. Then your super's calling because they don't know their ass from a hole in the ground."

He also told me that most of the bid status tracking lives in people's heads. When I asked where the system of record is for which bids are in, whose court it's in, what's missing, he said: "But like most of that's just living in our head."

That's a senior estimator spending his time on phone tag, email chains, and mental bookkeeping. The scope gap knowledge, the trade-splitting judgment, the ability to look at a set of bids and know something is off? That's maybe 20% of his day. The other 80% is logistics.

The Counter-Argument (And Why It Falls Short)

Now, some people will push back on the "estimator crisis" framing, and they have a point.

The BLS actually projects cost estimator employment to decline 4% over the next decade. Their reasoning is that estimating software is making each person more productive, so fewer are needed. Tools like Togal AI claim 97% accuracy on takeoffs completed in seconds versus weeks. One case study showed estimators doubling their bidding capacity after adopting AI-assisted takeoffs.

And the pipeline isn't as dry as the headlines suggest. Youth apprenticeships in construction grew 113% between 2010 and 2020. Construction management degree programs are expanding. Young people are entering the field.

So is this actually a crisis?

I think the productivity argument misses something important. Yes, AI tools can speed up a takeoff. Yes, fewer estimators might be needed for the mechanical extraction work. But the BLS is measuring headcount, not knowledge. Headcount isn't the problem. The problem is whether the estimators you have left can make the same quality decisions as the people they're replacing.

75% of construction companies haven't moved beyond initial conversations about AI. So you've got a workforce aging out, a technology that could help bridge the gap, and an adoption rate that's nowhere close to keeping pace with the retirements. The math doesn't work if you're losing experienced people faster than you're equipping the remaining ones.

What Actually Needs to Happen

I talked to another estimator who works public sector projects. He doesn't use scope sheets at all. His entire process is: look through the plans, call each sub individually, talk through scope on the phone, and make notes on their proposals during the conversation. No documentation system. No formalized workflow. Phone calls and memory.

When I asked about bid leveling tools, he said he tried one inside Procore and found it "hours upon hours of just setting up how to compare bids." He abandoned it after two attempts. His workaround is that he already knows his subs' tendencies. He knows who usually misses part A and who always includes items B and C without being asked.

That is irreplaceable knowledge. But it's only irreplaceable because nobody has captured it.

The experienced estimators I've talked to aren't opposed to AI. What they don't want is a generic tool that doesn't understand their process. One PM told me flat out: "I think we'll lose a lot if we take out the personal communication." A Gen X preconstruction lead said he hates AI-generated emails because they feel impersonal and spammy. These are real objections, and they're not wrong.

Those objections are about replacing judgment. Nobody is arguing for that. But 120 hours per project on subcontractor outreach and follow-up? That's not judgment. That's logistics. And every hour a senior estimator spends on phone tag is an hour their actual expertise, scope gap identification, bid analysis, trade-splitting decisions, sits unused.

The better version of this looks like systems that learn from how your best estimators actually work. What trades does your team typically split, and how? What scope items get missed most often on your project types? What does your specific scope sheet format look like? Which subs in your market are reliable for which kinds of work? That knowledge is configurable. It can be documented, structured, and built into workflows that survive after the person who created them moves on.

This Isn't a Crisis. It's a Transition.

The industry has always worked this way. Experienced people carry the knowledge. New people learn by sitting next to them for years. Eventually the experienced people move on and the cycle repeats. That's not broken. It's just slow, and it doesn't have to be.

Construction estimating still requires human judgment. Nothing I've seen in AI changes that. But the 197 hours of busy work per project, the phone tag, the manual scope sheet assembly, the mental bookkeeping? That stuff can go. And when it does, the experienced estimators on your team get to spend more time doing the work that actually requires their experience. The junior people ramp faster because the process is documented instead of improvised. And the knowledge your best people spent decades building doesn't have to live exclusively in their heads anymore.

Nobody needs to panic about this. But the firms that start thinking about it now, while their experienced people are still around to inform the process, are going to be in a better spot than the ones who wait until after the retirement party to realize nobody wrote anything down.