Where to start

Finding your footing when everyone else seems to already be running

A note before we begin: Carl and Dani are fictional — a constructed scenario, not a reported story. The failure mode they illustrate is real.

Carl has been in commercial construction for 26 years. He started as a laborer, learned estimating from a man who did it by hand on graph paper, and eventually bought the company. He’s got 31 employees, a solid reputation in mid-sized institutional work, schools and medical clinics mostly, and a backlog that used to feel comfortable.

Lately it feels thin.

He’s lost 4 bids in the last 8 months to a competitor he used to beat reliably. He doesn’t know exactly why. The scuttlebutt at the Associated General Contractors chapter meetings is that Hendricks Construction has gotten faster — faster estimates, faster responses, tighter proposals. Someone mentioned AI. Carl nodded like he knew what that meant.

Created with ChatGPT

The noise problem

Carl did his homework. He’s watched the videos, sat through 2 demos, read maybe a dozen articles. Every one of them seemed to be selling something. Every tool promised to change everything. None of them started with his actual day.

So he kept waiting. For clarity. For the right moment. For someone to just tell him where to begin.

His project manager, Dani, is 34. She came over from a larger regional firm 2 years ago, where she’d seen AI used in a few practical ways — nothing dramatic, but real. She’d been patient with Carl’s hesitation, more patient than she felt. But after the fourth lost bid she decided to say something.

She didn’t walk into his office with a tool recommendation. She walked in with a question.

“What’s the thing you hate doing most that you have to do anyway?”

Carl didn’t hesitate. “Owner updates. Writing up the weekly project reports. We’re good at building things. We’re not good at explaining what we built to someone in a suit who needs to feel informed.”

Dani had expected estimating. But this was better.

The right starting point

The weekly owner report at Carl’s company was a genuinely painful thing to produce. His superintendents were already logging everything in the Procore system— daily progress, open RFIs, inspection results, percent complete by trade. The data existed. But someone still had to pull it together into a coherent narrative, and that someone was usually Carl or his office manager, neither of whom wanted to be doing it.

Dani had seen this problem solved before. Carl’s firm was already using Procore. The data was already there. An AI tool connected to it could draft the owner report automatically, pulling what the supers had already entered. Carl would review the draft, spot-check anything that needs context, and send it. The client would get something that looks like it came from a firm twice Carl’s size.

“Your supers wouldn’t need to do anything extra,” she told him. “They’re already putting everything in Procore. You just won’t be spending 3 hours a week turning it into sentences.”

That landed. Carl understood that.

He didn’t need to understand large language models. He didn’t need a vendor relationship or an IT department or a six-month implementation plan. He needed one AI tool with a Procore connection and one super willing to let it run for a month.

How it played out

They started with Caleb, his most organized superintendent, on a 9-month school renovation in the suburbs. Caleb was already keeping Procore current — that was just how he ran his jobs. Dani set up the AI connection, configured the report template, and ran it the first Friday. It pulled Caleb’s logs, RFIs, inspection records, and schedule data and turned them into a 2-page owner report. Caleb read it, fixed 2 things the first week, 1 thing the second, and by the third week there wasn’t anything for Caleb to fix — just review and send.

The owner’s project coordinator called Carl after week 4. She said the reports were excellent. She asked if they’d changed something.

Carl said they’d been working on their communication processes. Which was true.

He felt a little sheepish about how long he’d waited. Not a lot. Just a little.

The theology underneath

Carl’s paralysis was not simple laziness. It was anxiety dressed up as caution — the fear of choosing badly, moving too soon, being fooled, wasting time, looking foolish. That fear is understandable.

Philippians 4:6–7 says “Be anxious for nothing, but in everything by prayer and supplication, with thanksgiving, let your requests be made known to God; and the peace of God, which surpasses all understanding, will guard your hearts and minds through Christ Jesus.” When we know that, as a good steward and wise leader we need to move forward, we should do it prayerfully, with thanksgiving, and trusting in God.

In GRACE terms, that’s an Allegiance question. What was actually driving Carl’s decisions? Fear of choosing wrong. Fear of being taken in. The unspoken hope that one more demo, one more article, would finally make the path clear enough to act. There’s a kind of information-gathering that is prudent — Scripture commends diligence, planning, counsel. But there’s another kind that is anxiety wearing wisdom’s clothes. At some point Carl wasn’t seeking clarity. He was seeking certainty God hadn’t promised, asking knowledge to do what only trust can do.

Dani’s question — what do you hate doing that you have to do anyway — was a way of quieting the noise. Stop looking at the tools. Look at the work. Find the place where the gap between what you’re doing and what you could be doing is concrete and small and fixable. Start there.

R.C. Sproul defines Biblical meekness as strength under control. Carl has genuine strength: 26 years of knowing how buildings go together, a crew that trusts him, a reputation he earned. What he needed wasn’t more information about AI. He needed to apply the disciplined approach he already knew — start small, prove the concept, expand from there — to a domain that felt unfamiliar.

A few questions to take back to your desk

Where in your business is the gap between what gets done and what gets communicated? Client-facing documentation is one version of this. There are others: internal handoffs, employee feedback, follow-up after service calls. The places where someone capable at the work is grinding through words they’d rather not write.

And who in your business might already have a sense of where to start? Dani wasn’t Carl’s technology officer. She was someone who’d seen more of the world than he had in one particular way, and she cared enough to ask a good question. That person might be in your company right now.

The tools matter less than you think. The problem you’re solving matters more. Find the friction first. The rest tends to follow.

This is actually what I do with small business owners — come in, ask Dani’s question a few different ways, and help identify where AI can do real work without requiring a major commitment to get started. If you’re in Carl’s position and aren’t sure where to begin, that’s a good reason to reach out.

I’ve helped businesses navigate a few of these waves — the Internet in the ’90s, smartphones in the ’00s. Each month I work with a small number of owners on exactly these questions around AI. If that’s you, reach out. I also publish a weekly AI market intelligence newsletter with a strategic and faith-based lens. Find it on Substack at Faithful Intelligence.

A note on process: This article was developed in collaboration with AI tools — reflecting the very intersection this newsletter explores. The research, theological convictions, and editorial judgment are my own; AI assisted in drafting and refinement.

A note on GRACE: The GRACE framework — Goods in Tension, Responsibility, Allegiance, Character, Effect — continues to run underneath these stories, even when it’s not named on the surface. In this article, you can find it in Carl’s anxiety about falling behind (Allegiance — what was actually driving him), in Dani’s instinct to start with the problem rather than the tool (Character — a formation habit, not just a productivity tip), and in who benefited when the right first step was taken (Effect — the supers, the owner’s team, the office manager). GRACE was never meant to be a checklist. It’s meant to become a habit of mind.