Dear Reader,

We’re shifting a few things on this newsletter. First, expect a new look. Minimal, black and white with shades of grey. Second, our Monday newsletter

Mondays - Before: Bricks, Bucks, Bytes episode announcement. Now: the executive briefing - weekly intelligence for construction leaders. Market shifts, thought leadership, and real insights from the ground.

You can read previous versions here: https://www.linkedin.com/newsletters/the-executive-briefing-7430824585270566912/

Wednesdays - No change, just a new look.

Fridays - Before: ICYMI Roundup. Now: A new look, with a condensed and more focused version capturing the 5 most important headlines from the week that matter… quality over quantity.

We’re looking forward to delivering you the most informed content in the construction industry.

Bricks & Bytes 💚

THIS WEEK
Platforms Went Shopping

Trimble and Autodesk both made acquisitions this week. For data, trust, and the workflows you already depend on. The operating layer of your business might have just shifted. Read on to learn more.

THE EXECUTIVE BRIEFING
THIS WEEK’S KEY TAKEAWAYS

Full episode write-up at the bottom↓

You can't vibe code trust. And the platforms know it. Trimble bought Document Crunch. Autodesk closed Rhumbix. Both this week. The platforms you already pay for are pulling in proven companies with real data and real customers. The switching cost just went up.

UK input costs hit their highest level since 1992. March PMI shows a 41-month high in input inflation. Nearly half of firms report rising costs. Work is coming in, but pricing hasn't caught up. Go check whether your bids from six months ago still hold.

The real AI return is in risk, not speed. A 31-year US contractor veteran splits AI into two buckets: efficiency and risk. His team's agent caught something a PM missed. Six-figure save. Most firms are putting their budget in the wrong bucket.

7 THINGS WORTH YOUR ATTENTION
ON THE RADAR THIS WEEK

Friday's March CPI could kill the rate-cut narrative — forecasters expect inflation to jump to 3.3% this spring.

ABC's Q1 construction forecast drops Wednesday — Anirban Basu on spending, tariffs, and recession odds. Free for members.

QXO closed its $2.25B acquisition of Kodiak Building Partners — 110 locations, $2.4B revenue. Brad Jacobs' roll-up is now live.

HS2 is lifting a 1,250-tonne viaduct section over the M6 this week — the biggest structural move of the year.

Data centre construction spending hit $25.4B in January alone. Full-year forecast is now $128B.

CRH's last day on the London Stock Exchange is April 17. If you hold shares through the LSE, move to NYSE before then.

Procore's next product launch wave lands April 21 — new portfolio analytics, estimating tools, and resource management incoming.

POWERED BY:

FULL EXECUTIVE BRIEFING
Platforms That Run Your Business Just Changed the Game

Two acquisitions landed in the same week. Different buyers. Different targets. Same signal.

Trimble bought Document Crunch, the AI platform that reads construction contracts and flags risk before it becomes a claim. Autodesk closed on Rhumbix, a twelve-year-old labour tracking company that gives contractors real-time visibility into workforce productivity. And if you think this is just a technology story, it's worth looking at what actually happened underneath.

Watch it HERE.

In both cases, a major incumbent platform identified a gap in its workflow, found a company that had spent years earning trust with real customers on real projects, and pulled it inside the walls. Not to bolt it on. To wire it into the core operating layer. Contract risk flowing directly into Trimble's project management and ERP systems. Labour data flowing into Autodesk's construction cloud. The PDF and the spreadsheet get replaced by something the system surfaces automatically.

This isn't new. Earlier this year, Procore acquired Datagrid. Late last year, Buildots acquired Genda. The pattern is the same every time.

The Buildots and Genda deal from late 2024 is worth understanding because it previewed exactly what happened this week. Buildots captures what has been built on site by comparing 360-degree camera footage to the BIM model. Genda tracks which workers from which trades are on site, where they are, and how long they spend in each location.

Separately, each data set tells you something useful. Together, they tell you something transformative. Not just that drywall is behind schedule, but that the crew has 60 percent of planned headcount and is losing 15 percent of its time waiting for the material hoist. Erez Dror, the Genda founder, described this as the difference between knowing what was built and understanding productivity. He said from the day Genda was founded, he showed every new employee a slide with Buildots and Genda on converging paths. He didn't know how it would happen. He knew it had to.

Rhumbix sits in the same territory. Labour timekeeping, productivity tracking, and unit rates captured at the point of work. Autodesk needed that data. They had design and document workflows. They had as-builts from the PlanGrid acquisition. But the workers, foremen, and payroll admins who live closest to the actual production were not in the ecosystem. Now they are.

Josh Levy from Document Crunch framed it as clearly as anyone. He said there are three moats that matter right now: product velocity, distribution, and data. And then a fourth that cannot be shortcut. Trust. You cannot vibe code trust. It has to be earned one customer at a time, one project at a time.

That fourth moat explains why these acquisitions keep happening instead of the incumbents building their own versions. Document Crunch has been at this for five years, pre-LLMs. Rhumbix has been at it for twelve. The machine learning, the customer relationships, the SOC 2 compliance, the enterprise-grade reliability, none of that can be replicated by spinning up a chatbot and calling it a product.

Dustin Devan, who built BuildingConnected and sold it for $275 million, was blunt about this on our show. He said there are founders right now vibe coding solutions and calling it a company. They have forgotten 70 to 80 percent of everything behind the scenes that makes software actually work in a deterministic industry where people get hurt if things go wrong.

ServiceTitan published a stat this week that ties it together. AI adoption among commercial contractors has more than doubled in a year, from 17 percent to 38 percent reporting measurable impact. But 59 percent of those adopters are using AI that is already built into software they use every day. They are not shopping for AI. They are buying workflows that happen to have AI baked in. The platform is the distribution mechanism.

While the platform layer consolidates, the financial pressure on the actual work is intensifying. UK March PMI data showed input cost inflation at a 41-month high, with nearly half of all firms reporting rising costs driven by energy prices and supply chain disruption.

Carl McFarland, a 31-year industry veteran who runs a major regional business for Big D Construction in the US, put the structural challenge plainly. Average margins in construction sit at two to three percent. Apple makes the entire annual profit of a $25 to $30 billion construction company in about 48 hours. That is the economic reality of this industry. And nobody has the balance sheet to run a 20-year transformation experiment.

Carl's point is that construction is living through its Blockbuster moment. Record profitability, especially in the AI infrastructure boom, masking the fact that the underlying business model is about to shift. Nobody rethinks their model during their best year. That is the trap.

The practical question for anyone running an open tender book: do the margins on bids won six months ago still hold given today's input costs? If the answer is uncertain, the conversation needs to happen now rather than at final account.

Carl also shared the clearest framework I've heard for how to think about AI deployment. His team separates AI work into two buckets. Efficiency and risk mitigation. The efficiency bucket is the buy-in work. Getting people comfortable. Helping a PM draft a report faster. That matters, but it is not where the serious returns live.

The risk bucket, Carl says, delivers ten times the return. He described an agent they built internally that collects monthly project metrics and surfaces them to the PM with insights. Human in the loop. Limited scope. The system caught something the PM had missed. That something had a six-figure number attached to it. The time saving from automating the report was three hours. The value of catching the risk was orders of magnitude higher.

MIT published a study this week that adds an uncomfortable layer to this. Their computer science lab showed that AI chatbots can push even perfectly rational users toward false beliefs through repeated agreement, a phenomenon they call delusional spiraling. The tool agrees with you, selects supporting evidence, and your confidence grows in something that may not be true. Even telling users the system might be biased did not eliminate the effect.

In a construction context, that means any AI tool reviewing risk-adjacent decisions, scheduling, cost forecasting, contract terms, needs governance around who reviews the output before it becomes an action. If nobody is checking the AI's answer, you do not have risk mitigation. You have risk amplification dressed in a dashboard.

The platforms are consolidating and they are getting smarter. The switching costs are going up. The firms that do not understand where their core platforms are heading in the next two years have a strategic gap that is widening by the quarter.

At the same time, the cost environment is tightening and the margin for error is shrinking. The intersection of those two forces, platform power and cost pressure, is where the next round of competitive separation will happen. The firms that treat their technology stack as a strategic asset and their AI governance as a leadership responsibility will pull ahead. The firms that treat both as someone else's problem will find out what the other side of that gap looks like.