M&A 7 min read

Attentive.ai Bought a Landscaping Takeoff Bureau That Was Already Building Its Own AI Tools

Announced from San Jose on 12 November, terms undisclosed. Takeoff Monkey calls itself 'a professional construction takeoff business', and its automation page — which opens 'we are construction professionals first' — lists AI document review, a schedule extractor and a retaining-wall scanner that compiles wall quantities. Attentive.ai paired the deal with four new trades in Beam AI's instant-takeoff workflow, and says 1,300-plus contractors have priced more than 500,000 jobs on the platform.

A curved paver pathway through a landscaped garden in low sunlight, bordered by a dry-stone retaining wall, clipped hedges and planting beds.

Attentive.ai announced three things in one release on 12 November, under a San Jose dateline distributed over BusinessWire and carried in full by the Rutland Herald: four new trades in the self-service workflow of its Beam AI takeoff product, in-app estimates for more trades, and what it calls its first acquisition. The target is Takeoff Monkey, a firm that sells construction takeoffs as a service and whose founder is listed on its team page in Winston-Salem, North Carolina. Terms were not disclosed.

The acquisition is the interesting part, and not because of its size, which is not disclosed. It is interesting because of what an AI takeoff company chose to buy.

What Attentive.ai Bought

Takeoff Monkey’s own site describes “professional takeoff services, enabling you to increase the quality and quantity of your bids while lowering your overhead,” and its about page is blunter still: “We are a professional construction takeoff business and help contractors of all sizes greatly enhance both the quality and quantity of bids.” The core deliverable is not software. It is a finished takeoff, and the company’s description of what comes with it is the most revealing sentence on the site: the delivered product is “full of notes pointing out discrepancies and contradictions within the documents, cost-savings ideas, detail references, and other critical details that estimators are looking for.”

That is estimator judgement, written down. It is the layer above measurement — the part where somebody who has priced this trade before notices that two drawings disagree, or that a detail reference points somewhere unhelpful. Takeoff Monkey’s founder Tommy Lather, whose about page says he has worked in the green construction industry for over 25 years, says as much in the release itself: “We built Takeoff Monkey because landscapers were losing bids to the clock. Doing takeoffs by hand helped our customers win more work, but every job still took hours of measuring on our end. Attentive.ai’s model does that in minutes.”

The company also builds software, and it reaches further than preconstruction plumbing. Its automation page opens with a direct positioning statement — “Most construction software is built by software companies. Takeoff Monkey is different. We are construction professionals first - estimators, operators, and workflow problem-solvers” — and the list below it runs from bid intake, folder creation and Monday.com job setup through addenda and revision tracking to a section headed “AI-Powered Document Review,” whose bullets include “Locate key details in drawings,” “Flag inconsistencies” and “Assist with QA review.” Its worked examples include an “Automated Schedule Extractor” that pulls “schedules, legends, and tabular data from PDFs directly into Excel” and a “Retaining Wall Scanner” that compiles “retaining wall elevations, TW/BW data, and wall quantities into organized spreadsheets.”

That matters for how the deal should be read. The target was not a firm that had stopped at measuring by hand; it was working toward the same layer from the other direction — estimators building tools for themselves in one trade, rather than software people generalising across many. So the asset reads as domain knowledge, a customer base and a set of trade-specific tools built by the people who use them, bought by a company that already has the general measurement model. Attentive.ai’s release says exactly that, in its own register: “Landscaping contractors have specialized takeoff and estimating requirements that require a deep understanding of their workflows and project needs. The acquisition allows Attentive.ai to build on Takeoff Monkey’s industry expertise as it expands its capabilities for landscaping professionals.”

The Deal and the Product Release Are the Same Move

The release pairs the acquisition with an expansion of trades in Beam AI’s “Do It Yourself” workflow, which it describes as giving contractors “instant, accurate AI takeoffs.” The release says four trades are being added, and names them as “Concrete-Rebar, Electrical, Landscaping & Paving.” Alongside that sits the second of the company’s two workflows, “Done For You,” which the release describes as delivering “QA-reviewed takeoffs and estimates within 24 to 72 hours for 15+ trades.”

Those two workflows are the structure worth noticing, because the acquisition could slot into both. A services bureau is, among other things, people who can staff the reviewed, human-checked path, and a body of completed work on a trade the instant path has only just started covering. Landscaping arrives in the DIY list and a landscaping-specialist services business arrives in the company on the same day. Whether the second is being used to train or quality-check the first is not something the release says, and this publication is not asserting that it is — only that putting the two announcements in one document invites the question.

Beam AI, the product, is published under its own domain and its FAQ notes it is “also known as iBeam”; its trade list there runs well past the four being added, covering flooring, demolition, insulation, snow, civil, utility and earthwork, painting, drywall, masonry, lumber and more. Attentive.ai’s corporate site lists two products — Beam AI for takeoffs and estimates, and BIM CoPilot for managed BIM services — and says Beam AI lets contractors “Bid 3X more jobs with accurate AI takeoffs & estimates,” with “99%+ accurate takeoffs & estimates in 24-72 hours.”

The Figures, and Whose They Are

The release’s headline numbers are all company-stated: Beam AI “is the platform behind $500 billion in construction bids”; “[m]ore than 1,300 contractors across the U.S. and Canada use it for takeoffs and estimates, and they have priced more than 500,000 jobs on the platform.” Co-founder and CEO Shiva Dhawan frames the strategy around capacity rather than accuracy: “Contractors need the capacity to pursue more work, but estimating teams can only take on so much when takeoffs and other preconstruction tasks require significant manual effort.”

Bid volume is the right metric for this product and a slippery one for a reader. A dollar figure for bids priced through a platform is a measure of throughput, not of won work, and it scales with the size of the jobs a customer happens to bid. It is a real number and it is not a revenue number, and the release does not present it as one.

The accuracy claim is the one with history. This publication examined the 99%-accuracy pitch across this category in June and found the reality more complicated than the marketing — AI counting is genuinely strong on clean, repetitive, well-drawn quantities and weaker everywhere the documents are ambiguous, which is most of the interesting cases. The benchmark picture is murkier still: two takeoff benchmarks published eleven weeks apart do not agree on how good frontier models are, with Togal.AI’s Construction’s Last Exam putting the best general-purpose model at 34.0 while a rival benchmark from Handoff put frontier models in the 40s and 50s. Vendor-reported accuracy for a trained, trade-specific product is a different quantity from a general model’s benchmark score, and none of these numbers are directly comparable. The point is only that “99%+” is a claim about a specific pipeline on specific document types, and it belongs to the vendor.

Where It Sits

The structural move — widen from the measurement outward, until a bid runs end to end inside one product — is now the common strategy in preconstruction rather than a distinctive one. Bobyard pushed its AI past the takeoff into the submittal in late October, so that a bid can run from drawings to priced estimate to branded proposal to submittal package in one place. PataBid took its first venture money after nine years to build a conversational estimating agent that ships in 2027. Attentive.ai’s version of the same widening is sideways rather than forward: more trades, and a trade specialist acquired rather than hired.

The release notes that the announcement “follows Convergence, Attentive.ai’s first conference for preconstruction professionals,” and the About section describes the company as trusted by “1,300+ leading businesses across the U.S. and Canada” and “rated 4.9 across G2, Capterra, and Software Advice.”

What Is Not Known

Terms were not disclosed, and no outlet this publication checked has a price. There is no stated headcount for Takeoff Monkey, no statement about whether the brand continues, and no indication of what happens to its existing services customers outside landscaping — its site advertises construction takeoffs, irrigation design and landscape design and drafting, while the release frames the deal almost entirely around landscaping. The release also gives no date for when the new DIY trades become available to existing customers beyond the present tense of “contractors can now get instant, AI-based takeoffs.”

The thing most worth watching is narrower than any of that. Takeoff Monkey’s selling point, in its own materials, was catching the contradictions in the drawings — in the notes on a delivered takeoff, and in the AI review tools it built to flag inconsistencies. The test is whether that survives the acquisition as a product feature, or whether it quietly becomes the thing the instant workflow does not do.