89% of CEOs say they'll hire more. Your estimator still doesn't exist.

It's 6:40 on a Thursday night and one light is still on in the office. It's your estimator. She has four bids due Monday, two of them for customers you've been chasing for a year. Right now she's copying line items out of a supplier PDF into the estimating spreadsheet. Then she'll copy the totals into the quote template. Then she'll check the freight tab that only she understands, because she built it.

You'd hire a second estimator tomorrow. You posted the job in the spring. You got résumés from people who had never priced a job, and one from someone who wanted the senior title and Fridays at home. The posting is still up.

The ceiling is a person you can't find

For a lot of growing companies, the next hire isn't a budget question. The money's there. The person isn't, at least not in your market, with the experience you need, this year.

That costs more than overtime. Bids you don't submit are revenue you never see, and nobody puts that on a report. Turnaround slips from two days to five, and the customer who needed a number by Wednesday buys from whoever answered first. Your best estimator gets tired, and tired people make pricing mistakes or update their LinkedIn.

The bigger cost shows up the next time you want to grow. New territory, new product line, a customer twice your usual size: every one of those runs through the same desk. Each growth step turns into another search for someone who doesn't exist yet.

Look closely at what she's doing at 6:40, though. Some of it is real estimating: reading scope, judging risk, knowing which supplier actually ships on time. A lot of it is moving numbers that already exist from one document to another. That second part is what's eating her week.

What fixing it actually looks like

The fix isn't a chatbot and it isn't a new estimating platform. It's a tool built around the way your shop already quotes.

When a request and the supplier quotes come in, the tool reads them, pulls the line items into your format, applies your pricing rules and that freight table, and drafts the quote in your template. Anything outside the normal range gets flagged: a margin that looks thin, a supplier price that jumped, a quantity that doesn't match the drawing. She reviews it, changes what needs changing, and sends it. The judgment stays with her. The re-keying is what goes.

Here's roughly how it gets built:

1. Sit with the person doing the work and map every step, every document, and every rule that lives in her head.
2. Build a working prototype and run it against a stack of real past quotes. Compare its drafts to what she actually sent.
3. If the prototype can't match her on old jobs, you find that out before paying for anything bigger. If it can, build the full version, connected to the systems you already run. You own the code.

Where it breaks: scanned drawings at bad resolution, handwritten field notes, and pricing rules that nobody ever wrote down and that change by customer. A good build flags those instead of guessing. You also need one person who owns the rules, or the tool drifts out of date the same way the freight tab would.

When I'd tell you not to build it:

• You quote ten jobs a month and every one is truly custom. A better template and an afternoon of cleanup will do more.
• Nobody agrees on the margin rules. Fix that first. Automating a messy process gets you a faster mess.
• Off-the-shelf estimating software already fits your trade and you just never finished setting it up. Finish setting it up.

The win is the same estimator getting more bids out the door and going home at six. And when you do find that second estimator, she walks into a job that's about judgment, not typing.

What the CEO survey says, and what it doesn't

KPMG's 2026 US CEO Outlook, released October 15, surveyed 400 US chief executives. 89% expect to increase headcount over the next three years, though most of them (60%) expect growth under 5%. 53% are also planning workforce reductions in some areas, and 86% are redesigning roles, career paths and team structure around human-AI collaboration. Their top workforce challenge is the skills gap, at 60%. Only 9% say AI agents have been transformational for their organization so far.

Two honest caveats. Every US company in that survey has more than $500 million in revenue (KPMG survey details), so these aren't your peers. And the reductions number is their plan, not the play I'm describing.

But the direction matters to you. The largest employers in your region plan to keep hiring, and they're rebuilding jobs around these tools at the same time. They're fishing in the same pond for experienced estimators, schedulers and coordinators. That pond isn't getting deeper. The companies that grow next year will be the ones whose existing people can absorb more volume without working until seven.

Try this before Friday

Pull the last quote that went out. Write down when the request came in and when the quote left. Then sit with your estimator for 30 minutes and walk through it together. Count every number she typed that already existed somewhere else: in a PDF, an email, another spreadsheet. Then split the elapsed time into hours someone was working on it and hours it sat waiting.

If the re-keyed count runs into the dozens and the turnaround runs into days, you have a capacity problem, not a hiring problem.

One next step

If you ran that and didn't like the number, book a 20-minute scoping call. We'll look at one process, and I'll tell you plainly whether it's worth building. If it isn't, I'll say so.