Ninety-one percent of contractors surveyed by Dodge Construction Network and CMiC in June 2026 reported tariff-related cost increases. That number, by itself, surprises nobody. Lumber duties on Canadian imports now total 45 percent. Kitchen cabinet tariffs doubled to 50 percent on January 1. Building materials costs have climbed 41.6 percent since December 2020, nearly double the 21.9 percent general inflation rate over the same period. Everybody knows.
Everybody in residential construction knows the squeeze is real, because the squeeze shows up in every bid, every change order, every phone call where a supplier says the price they quoted last Tuesday is no longer the price.
What the survey revealed that should unsettle anyone watching this industry: only about a third of contractors are even aware that AI tools for construction exist.
Two Industries Wearing the Same Hard Hat
Split the Dodge/CMiC results into two populations and the picture sharpens. Population A knows AI tools exist and, in many cases, has adopted them. Automated proposal generation scores 92 percent effectiveness among users, AI-driven contract risk review hits 86 percent, and progress tracking through automated site photo analysis scores 92 percent, meaning these contractors are compressing estimating timelines, catching risk clauses before they sign, and flagging schedule deviations from a photograph instead of a three-hour site walk.
Population B, roughly twice the size of Population A, is absorbing that same tariff hit, that same 41.6 percent materials inflation, that same labor squeeze, with the same tools they used in 2019.
The NAHB pegged the average tariff cost at $10,900 per home in its April 2025 Housing Market Index survey, a number that predates the Canadian lumber duty escalation to 45 percent and the kitchen cabinet doubling, which means the current figure is almost certainly higher. A small builder completing five homes per year faces roughly $54,500 in annual margin destruction from tariffs alone, layered on top of a construction cost base that was already 41.6 percent above 2020 levels before anyone in Washington picked up a pen.
Population A has options. Population B has a spreadsheet and a prayer. Same industry. Widening gap.
The Breakeven Nobody Calculated
Consider the economics for a small residential GC running $2 to $5 million in annual revenue. AI-powered estimating platforms like Buildxact, ProEst, and Togal.AI cost $200 to $500 per month. An estimator spending 40 hours per month on proposals, at a loaded cost of $45 per hour, represents $1,800 in monthly labor. If AI cuts that time in half, consistent with the Dodge/CMiC survey's 92 percent proposal-generation effectiveness rating, the savings run $900 per month against a $200 to $500 subscription. Payback on month one, no capital expenditure, no IT department required.
Scale that across the full tariff exposure: five homes at $10,900 each is $54,500 in annual cost pressure. AI estimating alone does not eliminate that number, but the downstream effects compound: fewer change orders (Mordor Intelligence reports a 41 percent reduction with predictive analytics), faster scheduling, earlier conflict detection. The gains stack.
Suffolk Construction, a top-ten domestic builder, attributes 40-plus hours of monthly time savings on a single multi-billion-dollar project to its AI-driven requisition process. Suffolk also spent $100 million over a decade building a 293-terabyte data lake to make that possible, which is approximately $99.7 million more than a five-home-a-year GC has available for technology investment.That cost asymmetry, the hundred-million-dollar data lake on one side and the five-hundred-dollar monthly subscription that most contractors have never heard of on the other, is the real policy problem, because when the government creates a cost shock through tariffs and the market creates a relief valve through technology, the question of who can access the valve becomes a question of industrial survival.
Who Can Afford to Adapt
Mordor Intelligence sizes the AI-in-construction market at $12.94 billion in 2026, growing to $27.92 billion by 2031 at a 16.62 percent compound annual rate. North America accounts for 42 percent of that revenue. General contractors represent 41 percent of AI expenditure.
Big numbers. Misleading ones, too. Comprehensive deployments, the kind Suffolk runs, frequently exceed $100,000 before producing measurable returns.
Residential subcontractors do not have $100,000 for technology experiments, and most residential GCs building fewer than twenty homes a year do not either. So they don't adopt. The affordable tier of AI tools, the $200-to-$500 monthly subscription platforms, solves part of the problem for firms that know the tools exist and have the bandwidth to evaluate, onboard, and integrate them into existing workflows. But the Dodge survey found that 65 percent of contractors report schedule impacts from skilled labor shortages and 76 percent cite retirements as a key driver of workforce attrition. The firms most devastated by the tariff-plus-labor squeeze are the same firms least likely to have someone on staff who can evaluate whether Togal.AI's takeoff automation is worth testing, because that person retired in 2023 and was never replaced.
Nobody is coming to fill that seat.
Seventy-four percent of respondents said they remain optimistic about their company's 2026 performance. That statistic should be read alongside the 84 percent who expect inflation to reduce overall construction volume this year, a pairing that reveals the quiet logic of survivorship: the firms confident enough to answer a survey about their outlook may not represent the firms that stopped answering surveys because they stopped operating.
What a Homebuyer Should Take From This
If your builder cannot explain what software tools they use for estimating, scheduling, and risk management, that silence has a cost. A real one. It sits in your bid already, a line item nobody broke out because nobody could explain what tool would have prevented it. A contractor absorbing $10,900 in tariff increases per home without any technological offset is passing that cost to you, building it into a thinner margin that increases the likelihood of corners cut during construction, or both. Ask. An honest answer that includes the name of an estimating platform, a scheduling tool, or a document management system tells you more about how your project will be run than any glossy brochure or Houzz review ever could.
If they look at you blankly, they may still be an excellent builder. Plenty of skilled contractors have thrived for decades on experience, relationships, and craftsmanship alone. But they are fighting a 41.6 percent materials cost increase, a 45 percent lumber tariff, a labor force that is retiring faster than it is being replaced, and they are doing it without tools that their more technologically literate competitors report make them 50 to 100 percent more effective. The math on that equation has an expiration date, and the tariff clock, the labor retirement clock, and the technology adoption clock are all running simultaneously whether any individual contractor chooses to acknowledge them or not.
Limitations
The Dodge/CMiC survey does not disclose its sample size, selection methodology, or whether respondents skew toward commercial or residential contractors. The "50 to 100 percent effectiveness" figures are self-reported by adopters, introducing both selection bias and measurement uncertainty. NAHB's $10,900 per-home tariff cost figure dates to April 2025 and may not reflect current tariff levels following subsequent policy changes. Suffolk's case study involves commercial megaprojects, not residential, and its efficiency gains may not translate linearly to smaller-scale homebuilding operations where the data infrastructure investment is orders of magnitude lower. The breakeven calculation above assumes a specific estimator cost, hourly rate, and time reduction that will vary by firm, market, and project type. Mordor Intelligence's market sizing relies on modeling rather than observed revenue data.