AI in the Scaffolding Industry

At the Scaffold & Access Industry Association annual convention in Nashville last month, I had the opportunity to attend a presentation by two good friends from Scaffold Resource, Gary Boncich, President and CEO, and Jim Vecchione, COO. Gary and Jim walked through how they’re using AI to operate their scaffolding business. It was well done, and more than that, it was thought-provoking.

They organized the discussion around four areas: safety, finance, operations, and sales, and showed both what Resource is doing today and where they believe the technology can go next. What stood out to me was not just what AI could mean for operating a scaffolding business more efficiently, but what those improvements could ultimately mean when it comes time to sell the company.

I have spent my career advising owners of industrial services businesses through that transition, so that is where my mind naturally went.

Construction has historically lagged many other industries when it comes to digitization and productivity improvements. In some ways, that creates an even bigger opportunity. Computer vision, connected data, and generative AI are moving beyond isolated pilot programs and becoming practical tools across safety, finance, operations, and sales.

For an owner thinking about the long term, these are more than operating improvements. Each one can help build a stronger, more valuable, and ultimately more sale-ready business.

Inventory: From Guesswork to Ground Truth

Ask a scaffolding company owner how much pipe, coupler, and plank they actually have in the yard versus what is on a job site, in transit, or lost to shrinkage, and the answer may not be as precise as you would expect. Physical inventory counts are labor-intensive, quickly become outdated, and may only be reconciled once or twice a year.

In a business where equipment represents significant working capital, that is a meaningful blind spot.

AI-powered tools are beginning to close that gap. Computer vision paired with drones, fixed cameras, or handheld scanning technology can count stacked components in a yard far more quickly than a traditional physical inventory. That same technology can also help reconcile equipment across job sites without requiring every individual piece to be tagged.

RFID at the individual-component level has been discussed in the industry for years, but deploying and maintaining it at scale can be difficult. Photo- and scan-based counting may prove to be a more practical near-term solution.

The result is not simply a better inventory count. It is the potential for a live, searchable asset ledger.

When that inventory data is connected to the financial system, it becomes much more than a logistics tool. Management can begin to:

• Analyze revenue by customer, segment, or region against the equipment deployed to generate it
• Evaluate margins by job type and identify which segments are driving profitability
• Track profitability job by job using actual equipment utilization and carrying costs, not simply labor and materials

Buyers want to understand which customers and job types actually make money. A company that can answer those questions with reliable data rather than management estimates demonstrates a level of operational sophistication that can matter during a sale process.

Profitability by Piece of Equipment

Job-level margins tell you which contracts are worth pursuing. Equipment-level profitability adds another layer.

Once inventory is tracked and connected to financial data, a company can begin asking more detailed questions:

• What’s this piece’s utilization rate — how many billable days a year is it actually out on a job versus sitting in the yard?
• What’s its all-in carrying cost — depreciation, maintenance, storage — against the revenue it’s generated over its life?
• Which SKUs or equipment classes are quietly underperforming, and which are prime candidates to not replace at end of life?

There is an important limitation to this analysis. Scaffolding is a systems business. A component that appears unprofitable on its own may still be necessary to complete a larger, profitable job.

The goal is not to eliminate every piece of equipment with a poor standalone return. The goal is to understand why it performs the way it does and how it contributes to the broader system.

That turns a large equipment fleet from a single capital expenditure category into a portfolio of assets that can be better understood and managed. From a diligence perspective, that information can be extremely valuable. A seller that can demonstrate equipment-level returns and explain the broader system economics can give a buyer much greater confidence in where margins are coming from and how the fleet is being managed.

Forecasting and Capital Investment Decisions

Once a company has reliable utilization and profitability data, AI can help turn historical information into a forecasting tool.

By combining equipment data with current job schedules, the sales pipeline, and historical seasonality, management can begin answering questions such as:

• Based on committed work and the probability-weighted pipeline, do we have enough equipment for next quarter?
• Where are utilization bottlenecks most likely to occur?
• Are those constraints concentrated in a particular component type, yard, or region?
• If we win several large jobs currently in the pipeline, what equipment shortfalls will we face?

Those forecasts can then inform capital decisions:

• Do we need to buy at all, or can pipeline timing be smoothed with rentals or inter-yard transfers?
• How much — right-sized to a realistic pipeline-weighted demand curve, not a round number picked because “we always buy a truckload in Q1”?
• What’s the payback — modeled against that specific equipment class’s historic utilization and margin, so the ROI case is grounded in the company’s own data rather than a vendor’s sales pitch?

There is also an important M&A question embedded in those purchases: is the capital expenditure maintenance or growth?

Buyers view the two very differently. Maintenance capex is the amount required to replace worn or obsolete equipment and sustain existing operations. Growth capex supports new customers, new markets, or additional volume.

A seller may characterize a purchase as growth capex, but a buyer will want evidence that the investment actually produced growth.

AI-supported utilization and revenue data can help establish that connection by tying equipment purchases to the incremental jobs, customers, or revenue they supported.

That turns a statement like “we bought more scaffolding” into a much stronger one: “we invested in additional scaffolding to support growth, and here is the revenue and return that investment generated.”

That is where AI-driven inventory management becomes a capital allocation story. A company that can demonstrate disciplined investment decisions and historical returns is presenting buyers and lenders with a much stronger financial story.

Finance: Putting AI Where the Numbers Live

One of the most practical parts of the Scaffold Resource presentation focused on finance.

The idea is not to ask a generic AI tool to interpret financial information in isolation. It is to connect AI to the systems the company already trusts.

Scaffold Resource described Sage as its authoritative accounting system, with CData Connect AI and Zapier providing the data and workflow layer and Claude serving as a natural-language interface for finance and operations analysis.

That can fundamentally change the workflow.

Instead of exporting CSV files, building pivot tables, comparing AR aging against project status across multiple spreadsheets, and making several calls to answer a relatively basic question, the finance team can increasingly interact directly with trusted company data.

The same concept can be applied to AP automation, anomaly detection, WIP analysis, project-level cash flow forecasting, and month-end reporting.

For an owner, the benefit is speed and visibility. For a buyer, the benefit is confidence that management’s answers can be traced back to the same source of truth the company uses to operate the business.

Scheduling: Matching the Right People to the Right Job

Scheduling in scaffolding is not simply a question of whether a crew is available. It is whether the right crew is available.

Workers need the proper certifications, experience, and skill set for each job. A complex industrial project may require a very different crew than a more straightforward commercial build.

Managing those variables through memory or spreadsheets becomes increasingly difficult as a company grows.

AI-assisted scheduling can help by:

• Qualification-aware assignment — cross-referencing each worker’s certifications (competent person, OSHA training, confined space, rigging, etc.) against what a specific job actually requires, so a crew never gets assigned without the credentials the site demands
• Skill/experience matching — pairing more complex or higher-risk builds (industrial coatings platforms, suspended scaffold, multi-story systems) with crews that have the right track record, rather than whoever happens to be free
• Dynamic re-optimization — when a job gets delayed, compressed, or a change order shifts scope, the system can re-shuffle crews and equipment across the broader job board in real time, instead of a scheduler manually reworking a whiteboard

The ability to dynamically adjust may be one of the biggest benefits.

Scaffolding schedules move constantly. Weather, customer changes, delays from other trades, and shifting project timelines can affect not only one job but every other project relying on the same labor and equipment pool.

AI-assisted scheduling can identify conflicts earlier and recommend ways to absorb changes without requiring someone to manually rebuild the entire schedule.

For a buyer, this matters because labor and scheduling issues are significant sources of margin erosion. Idle crews, improper certifications, and mismatched experience can quickly affect profitability and increase risk.

A company with a disciplined, systematized scheduling process is also less dependent on the institutional knowledge of one key employee.

Closing the Field-to-Office Loop

Gary and Jim also focused on a problem every contractor recognizes: valuable information dies on the way back to the office. Daily reports get hand-typed, photos stay buried on phones, decisions live in text or WhatsApp threads, and important context exists only in a phone call or in someone’s head.

AI can turn a foreman’s voice note into a structured daily report tagged to the right job, route site photos into a searchable project record, make live project information conversational, and identify recurring issues across jobs before they become write-offs. That is not really an app story. It is a field-to-office information story, shortening the time between something happening on a job and the rest of the organization knowing about it.

Safety: More Visibility Across Every Job

Scaffolding safety has traditionally relied heavily on periodic human inspection. A competent person walks the structure, supervisors conduct visual checks, and incidents are documented when something goes wrong. That model is only as good as the last inspection; a lot can change on a live job site between walkthroughs. AI-enabled wearables and drones start to close that gap by making inspection continuous rather than periodic.

• Wearables on crew members can monitor things like fall-arrest engagement, proximity to hazards, fatigue indicators, or unsafe posture and lifting patterns, flagging issues to a supervisor in the moment rather than surfacing them in a post-incident review
• Drones with computer vision can fly a structure and check tie-ins, bracing, plank condition, and load consistency against the engineered plan — catching a missing guardrail or an out-of-spec configuration without waiting for a person to physically climb and check
• Hazard detection can extend beyond the scaffold itself — spotting weather-related risk (wind loading, ice), site congestion, or other trades encroaching on the structure

Scaffold Resource showed a simple example using daily JHAs: ask the system for the most common flagged responses over the prior 90 days, and management can immediately see recurring issues such as first aid kit availability, Stretch & Flex, 911 functionality, and evacuation-location planning, and then target corrective action. Layered over time with inspection, incident, near-miss, and weather data, that same approach can help identify patterns and jobs that deserve earlier attention. AI does not replace the competent person or safety professional; it gives them more data, fewer blind spots, and an audit trail that is built as the work happens.

From a risk and valuation standpoint, this may be the single most important area in this whole discussion. Safety incidents are the fastest way a scaffolding company’s insurance costs, bonding capacity, and reputation take a hit — and by extension, its multiple. A company that can point to real-time, AI-monitored safety compliance, with a documented trail on customer-caused modifications, is de-risking the exact liability that keeps buyers, insurers, and sureties up at night.

HR Compliance: I-9s as a Deal Risk

For most industries, I-9 compliance is a back-office formality. In scaffolding and industrial services more broadly, it has become one of the more common landmines in M&A diligence. Labor-intensive, high-turnover, multi-site workforces make manual I-9 administration error-prone: missed reverification dates, inconsistent documentation across job sites, forms completed late or filled out incorrectly. Any one of those is a compliance exposure; across hundreds of employees over multiple years, it becomes a pattern a buyer’s counsel will find.

AI-assisted I-9 review can get ahead of that:

• Document-level checks — scanning submitted I-9s for missing fields, expired documents, mismatched dates, or signatures out of sequence, the kind of errors that are easy to miss at the volume a scaffolding company processes
• Reverification tracking — flagging work-authorization expirations before they lapse, rather than discovering a gap during an audit or a diligence request
• Pattern detection across sites — surfacing whether a particular yard or supervisor has a disproportionate error rate, pointing to a training gap rather than random noise
• Audit-readiness scoring — giving ownership a running sense of how clean the full I-9 file would look if it were requested tomorrow

This one hits close to home in my seat, since I’m often the one delivering news of an I-9 problem to a seller mid-process or defending a company’s file to a nervous buyer. A scaffolding business that can show clean, continuously monitored I-9 compliance removes a diligence item that too often turns into a price chip, an escrow holdback, or in a bad case, a deal-killer. It’s a good example of AI providing not just efficiency, but genuine risk transfer at the negotiating table.

Legal Review: Faster, More Consistent Contract Review

Scaffolding companies rarely get to dictate contract terms; they’re typically handed a customer’s or general contractor’s master service agreement and told to sign it to get on the job. The problem is those agreements are often dense, inconsistent from customer to customer, and loaded with terms that quietly shift risk onto the scaffolding company: uncapped indemnification, onerous insurance and additional-insured requirements, pay-if-paid clauses, unfavorable warranty periods, or liability provisions that go well beyond what the work itself justifies. Getting real legal review on every one of these, on the timeline a job demands, is expensive and often just doesn’t happen — someone signs it to keep the job moving.

AI-assisted contract review can change that turnaround:

• Clause-level flagging — scanning an incoming agreement against a library of the company’s known “acceptable,” “negotiable,” and “walk-away” terms, so a project manager gets a fast read without waiting days for outside counsel
• Indemnification and insurance checks — specifically testing whether the risk-shifting language and required coverage limits line up with what the company’s own policies can actually support
• Redline suggestions — proposing specific alternate language for the terms that fall in the “negotiable” zone, so the back-and-forth with the customer’s legal team moves faster
• Pattern tracking across customers — building a running picture of which customers consistently push the most onerous terms, useful both operationally and at the negotiating table
• Negotiation-history recalls — surfacing what a specific customer has previously agreed to accept, so the next negotiation doesn’t start from a blank page. If a customer already negotiated their indemnification cap or insurance terms on a prior job, that precedent becomes the starting point for the next agreement, rather than the company re-fighting the same battle from scratch every time paperwork comes in

The real value is consistency at speed. A scaffolding company signs dozens or hundreds of these agreements a year across different customers and job sites; AI review means the same standard gets applied every time, instead of it depending on which project manager is under the most time pressure that week. It also means negotiation doesn’t have to start over with every new contract — pulling up what a customer has already agreed to in the past turns a lengthy back-and-forth into a quick confirmation. And it builds an audit trail that is useful in a dispute, and useful in diligence, where a buyer wants to know the company hasn’t quietly accumulated a portfolio of contracts with unlimited exposure. Worth noting: this is a great efficiency story, but AI review should augment legal judgment on higher-risk agreements, not replace it. A buyer or insurer will want to know a human still signs off on anything outside the pre-approved parameters.

Bidding: Pricing Jobs on Data, Not Instinct

Bidding is where a lot of scaffolding companies quietly give away their margin or price themselves out of work because the estimate leans heavily on the experience of whoever’s building it that week. Every job has its own mix of variables — structure complexity, site access, customer difficulty, duration, region — and it’s hard for any one estimator to hold years of historical outcomes in their head while pricing a bid due tomorrow.

AI-assisted bidding puts that history to work directly:

• Similar-job matching — pulling comparable past jobs by customer, segment, structure type, or scope, and showing what was bid versus what it actually cost to deliver
• Customer-specific patterns — flagging that a particular customer historically runs long, generates more change orders, or squeezes margin, so the bid prices in that reality instead of assuming a clean job
• Margin-at-risk modeling — surfacing which inputs (labor hours, equipment days, site conditions) have historically been the most likely to blow past estimate, so the bid builds in the right contingency instead of a flat, one-size-fits-all buffer
• Win-rate feedback loop — tracking bid-to-win ratios by price point and job type, helping the company see where it’s underpricing to win work versus leaving profitable jobs on the table

The Scaffold Resource presentation broadened that same idea across sales and estimating: compare new scopes and schedules against historical jobs; review proposals side-by-side for pricing, scope, and exclusions; pull component counts and material availability from the ERP while the job is being priced; and use generative AI to accelerate proposals, RFP responses, and follow-up communication. The point is not simply to write faster. It is to give the estimator and salesperson better access to the company’s own operating history while the decision is being made.

The real shift here is turning every completed job into training data for the next bid, rather than letting that knowledge live in one estimator’s head and walk out the door when they leave. It also gives ownership a clean answer to a question buyers care about: is this company’s revenue growth coming from disciplined, profitable bidding, or from winning work at thin (or negative) margins that won’t hold up post-close?

Closing the Gap Between What’s Sold and What’s Built

Beyond pricing the job right, there’s a second, quieter risk in bidding: the handoff between sales and operations.

A salesperson can win a job based on a scope, schedule, or set of assumptions that never gets fully translated to the crew responsible for actually building it. Maybe site access is different than expected. The structure is more complex than what was priced. The schedule is tighter than the field team ever agreed to.

None of those issues necessarily make the original bid look bad on paper. The problem shows up later, when a job that “should have been profitable” isn’t.

Because at the end of the day, what was sold and what had to be delivered were two different jobs.

AI can help close that gap directly:

• Scope-to-spec reconciliation — comparing the sold proposal (drawings, scope narrative, assumptions) against the engineered plan operations is actually building to, flagging mismatches before crews mobilize
• Assumption tracking — surfacing the specific site conditions, access, or schedule assumptions baked into the bid, so operations knows exactly what the price was built on and can flag early if reality doesn’t match
• Change-order triggers — automatically detecting when actual conditions deviate from the bid’s assumptions, so a change order gets raised in real time instead of the difference just being absorbed as margin erosion
• Sales-ops feedback loop — feeding completed-job data back to sales, showing where sold scope and delivered scope diverged, so future bids get tighter and salespeople get better calibrated over time

Pricing the job right only matters if what gets built matches what got priced. A buyer diligencing a scaffolding company will want to know that a strong bid-to-win ratio actually converts into delivered margin — and this is the mechanism that protects that conversion.

What Scaffold Resource Is Doing Today and What Comes Next

What made Gary and Jim’s presentation useful was that it was not framed as a distant-future exercise. Scaffold Resource described a deliberately small stack: Sage as the system of record; CData Connect AI and Zapier as the data and workflow layer; Claude as an AI assistant for finance and operations analysis; and people and processes as the final workflow layer. They are also piloting AI-assisted safety and inspection documentation, with connected forecasting and other use cases on the roadmap.

Their lessons so far were straightforward: data plumbing first, AI second; start where the pain is real; keep humans in the loop on purpose; a small, connected stack beats a collection of disconnected shiny tools; and measure the first wins in hours returned to operators before trying to force a dollar ROI calculation.

The roadmap moves from conversational financial reporting, voice-to-daily-report workflows, and photo-based inspection logs toward historical estimating, materials and yard forecasting, PPE computer vision, project cash-flow forecasting, scheduling and dispatch support, predictive risk scoring, proposal-versus-contract review, scope comparison against historical jobs, and instant material lookups during bids. That progression matters: the more sophisticated applications depend on the quality of the data and workflows underneath them.

There are real risks, too. Bad source data produces confident-sounding bad answers. Customer, crew, and financial information need clear confidentiality rules. Companies need to avoid over-reliance on outputs, design systems they can change as vendors evolve, and earn workforce trust so employees see AI as a tool that helps them rather than a stopwatch pointed at them. And on safety-critical, legal, and business-critical decisions, human oversight remains essential.

The Real Dependency: Data In, Value Out

Everything above is only as good as the data feeding it. AI doesn’t manufacture insight out of nothing — it’s dependent on staff in the yard, on the job site, and in the office consistently and accurately entering the information that trains and feeds these systems. A scan that doesn’t get done, a job that doesn’t get logged, a certification that doesn’t get updated — any of these leaves a gap, and gaps compound across a system like this.

That makes this as much a culture shift as a technology rollout. It requires real buy-in from the people doing the day-to-day work, many of whom have run things their own way for years and have no obvious reason to change unless they understand what’s in it for them. The case worth making to employees isn’t abstract — it’s that better data leads directly to a better-run, more efficient, more profitable company: fewer scrambles for missing equipment, schedules that actually hold, bids that don’t quietly lose money, a business that’s more stable and easier to work for. When staff can see that connection, consistent data entry stops feeling like extra paperwork and starts feeling like ownership in the outcome.

Bringing It Together: Sale Readiness as the Payoff

Every area above — inventory, equipment profitability, forecasting, scheduling, safety, HR compliance, contract risk, bidding — generates data. The real payoff comes when a company keeps that data organized and ready, not scrambled together the month an LOI shows up. That’s the difference between a smooth process and a bruising one.

We saw this play out recently in the sale of CD Specialty. Our client came to the table with clean, well-organized financial and equipment data, the kind of documentation most sellers say they have and then can’t quite produce when a buyer’s diligence team actually asks. That preparation didn’t just speed up the process; it protected the number. There was no scrambling to reconstruct equipment records, no gaps that gave the buyer’s team an opening to chip at the price, no unanswered questions that erode confidence and, with it, leverage.

That’s the throughline for this whole article: AI isn’t just an operating efficiency play for scaffolding companies day-to-day. Across the same four anchors Gary and Jim laid out — safety, finance, operations, and sales — it can become a running diligence file being built in the background, year after year. A company that uses AI to track inventory and equipment ROI, forecast demand, schedule qualified labor, capture field information, monitor safety trends, keep I-9s clean, manage contract risk, and bid profitably isn’t just running a tighter business. It’s building a data room in real time, so that whenever the moment comes to sell, the answer to every hard buyer question is already sitting there — organized, defensible, and ready to go.

For an owner, that’s the real argument for adopting these tools now: it’s not just about running the business better today. It’s about making sure that when it’s time to sell, the value you built shows up fully in the price.