Every estimator has lived some version of this evening. A family member calls, thrilled, phone held up to the camera to show off the AI generated rendering their design app spat out for the kitchen remodel, moody pendant lighting and all, produced in about the time it takes to say "open concept." Meanwhile the estimator is on their fourth cup of coffee, cross referencing addendum three against the electrical scope, because the drawing set contradicts itself and nobody on the design team can explain why. The family member asks, not unkindly, why the actual bid can't move at the same speed as the pretty picture. The estimator does not have a great answer, mostly because it is a fair question.

That gap, between how fast AI moves on the glossy end of the industry and how slow it still moves on the arithmetic end, is not just a dinner table observation anymore. A survey released this week put real numbers on exactly where AI adoption is landing in construction and design, and the pattern lines up with that phone call a little too well.

Houzz published its 2026 U.S. State of AI in Construction and Design Report this week, based on a survey of 601 construction and design businesses fielded in June and July, paired with a companion survey of 3,149 U.S. renovating homeowners fielded in June. The headline number is that 41 percent of pros now use AI for everyday business tasks, up from 34 percent in 2025. Split by trade, construction firms pulled ahead sharply: 52 percent now use AI regularly, up 20 percentage points year over year, compared with 39 percent of design firms, up 8 points. That is a real reversal of the assumption that AI adoption would start with the creative side of the industry and trickle down to the builders. This year the builders moved faster.

The productivity numbers behind that adoption are the more interesting half of the report. Fifty two percent of pros using AI report saving three or more hours a week, and 18 percent save eight or more, an entire workday returned every week. Houzz translated that into dollar terms and landed on an average of 244,000 dollars in annual productivity benefit per construction firm using AI, against a smaller but still real figure for design firms, who save roughly 3.6 hours a week on average. Numbers like that are easy to wave off as a vendor sponsored survey doing vendor sponsored survey things, but the consistency across firm size and trade is what makes it worth reading past the headline.

Where that time is actually going is the part worth sitting with. The top business functions for AI use were sales and marketing at 64 percent, planning and design at 61 percent, project and client management at 59 percent, administrative work at 52 percent, and business operations at 45 percent. That ordering is not an accident. AI is winning first at the tasks that produce something a client can see immediately: a rendering, a mood board, a proposal deck. The tasks that produce a number a client can hold you to, a quantity takeoff, a scope reconciliation, a bid that actually matches the drawing set, sit further down that list and further behind in tooling. That is the exact corner of the industry we spend our time in at BidForgeAI, and it is not a coincidence that the report's own ranking shows sales and marketing AI outpacing the harder, less visible math that decides whether a project makes money at all.

The homeowner half of the survey backs up the same story from the buyer's side. Twenty two percent of renovating homeowners said they had used AI somewhere in their own project, skewing heavily young, 41 percent of Millennials versus 7 percent of Seniors. Among homeowners whose pro used AI, the benefit they cited most was clarity: 76 percent said it helped them visualize the project, 49 percent said it helped them understand their pro's ideas. Homeowners are not asking AI to reconcile a spec sheet. They are asking it to show them what the kitchen will look like, which is precisely the fast, visual, low stakes use case where the technology already performs well.

None of this is uncomplicated enthusiasm, which is the part that keeps the report honest. Concern about AI reliability and accuracy actually grew alongside adoption, from 33 percent of pros in 2025 to 37 percent this year, even as 97 percent of respondents said they expect AI to meaningfully reshape the industry within five years. That is not a contradiction so much as a fairly accurate read on where things stand: people trust AI to help them pitch and visualize a project faster than they trust it to get the underlying numbers right, and this year's data suggests that trust gap is widening even as usage climbs. The rendering will keep getting faster. Whether the bid catches up is the more interesting question, and it is the one worth watching this survey answer differently next September.

Sources