We lose deals to fit. Not to price, not to a competitor, to the project simply not being the kind of project this helps on.
That's worth writing down, because a demo can hide it. You'll watch a scope package get built in under an hour, decide it's impressive, run it on your next pursuit and get much less out of it than the demo suggested. The tool didn't underperform. The project was wrong for it.
Here's how to tell before you spend anything.
AI scope review pays off when three things are true at once.
You're writing the scopes. The value is in turning a document set into trade packages your subs price. If your team authors the scopes of work, that's the workflow.
The document set is reasonably complete. Drawings, specs, addenda, enough that reading them against each other is the actual labor. Construction management delivery tends to fit well for this reason. A senior PM at a Canadian ICI GC described the pressure in the playbook interviews: "Our construction management clients expect us to find the scope gaps in the design too now. They expect us to be designers and engineers."
Somebody feels the pain. There's a person on your team spending 30 hours a bid on scope packages. Without that person, nothing changes, because nobody's hours come back.
When it fits, it really fits. EllisDon used Scope Agent output to issue a formal addendum modifying subcontractor scopes of work, which as far as we know is the first documented case of AI-generated scope flowing straight into a bid package at a top-tier GC.
Four situations where we'd tell you to wait. All four came from real calls this summer.
One customer couldn't use it on his current project because the prime contractor had authored every scope document. He was handing those to lower-tier subs to bid. There's nothing to generate. Reviewing what you were handed is a different job, and a smaller one.
A preconstruction director told us plainly that his recent pursuits were design-build or small, so running them through wasn't worth it yet. He's right. On design-build the design is moving under you, and scope qualification against a set that changes weekly means re-running against a moving target. The value shows up later, once the set stabilizes.
One prospect said the contractor view wasn't a tool he'd need. Fair. If your output is an internal estimate rather than packages a sub prices, you're paying for the half of the product you won't use.
Three separate teams asked us for this in twelve days: give me a report of what changed between issued-for-tender and issued-for-construction. One wanted it for GMP reconciliation.
We don't do it. Provision reviews the current set. It won't diff one set against another and hand you a change report, and no amount of clever prompting gets you there. If that's your problem, this isn't your tool.
Notice what isn't on either list: your revenue, your headcount, your project value.
Fit here is about documents and workflow. A team running one complex hospital project a year can get more out of this than a team running twenty small tenant fit-outs, because the hospital has the document volume where scope gaps hide. Project count is a bad proxy. Document complexity is the real one.
Delivery method matters more than either. Construction management and design-bid-build, where scopes get authored against a defined set, fit better than design-build in its early phases.
Run it on a project you already closed.
Pick one where you know where the gaps turned up, because you paid for them. Feed in the set as it existed at bid. Then check whether the output names the items that later became change orders.
You get a scored answer in an afternoon instead of an opinion. If it catches what you missed, the business case writes itself. If it doesn't, you've learned that on a dead project rather than a live one.
That test costs you a couple of hours and it's the one we'd run if we were you. Bring a closed project and we'll run it with you. If you want the mechanics first, Scope Agent covers what it produces, and Risk Review covers contracts and specs.
It works better once the design set stabilizes. In early design-build phases the set moves too much for scope qualification to hold, so we'd suggest waiting.
Then there's less to gain. The product generates scope packages from a document set. If you're passing down scopes somebody else authored, that workflow isn't the one being automated.
No. It reviews the current set and won't produce a change report between two issues of the drawings.
No. Fit tracks document complexity rather than project value. A document-heavy project benefits more than a simple one of the same contract value.
Construction management and design-bid-build, where your team authors scopes against a defined document set.
Run a project you've already finished and check the output against the change orders you paid.
Run Scope Agent against a finished job and check the output against the change orders you paid.
Book a WalkthroughMore Articles
