The strongest thing a tool can tell you is when not to use it. This page is that list. It's the cases where BidCertain — or any bid-intelligence / AI estimating tool — is not the right way to get the number, and where your own estimator's judgment, your PE, or a simpler tool is the better answer.
We're publishing this because honest "when to pass" framing is exactly what a competent estimator expects and exactly what earns trust. A tool that tells you every problem is a nail it fits is not a tool you can rely on. This list is the other side of the same screening discipline that flags a WEAK bid-fit or an owner you can't get paid by — the structure that tells you when to walk away applies to the tool itself.
When AI estimating is not the tool
1. When quantities are the whole job and the set is simple
If the job is a small, well-scoped set where the takeoff is the estimate — a few line items, a clean existing condition, no ambiguous interfaces — then a drawing-measurement takeoff tool or a spreadsheet is the right answer. Bid intelligence isn't a takeoff engine, and it shouldn't pretend to be. For a clean, small, repeatable scope, measure it directly and be done.
2. When the answer needs a licensed engineer's seal and you don't have review arranged
Any design output that touches structural capacity — a pile, a footing, a retaining wall, a shoring system — requires review by a licensed professional engineer before you rely on it. If you don't have a PE or qualified reviewer in the loop, an AI design-assist pass is not the tool: the preliminary sizing it proposes has no path to become a sealed, bidable design. That's a process problem, not a software gap — but it means the tool can't help until you solve the review.
3. When the ground truth is physically uncertain and untested
The CMAA go/no-go frame treats "understand the risk" as a gate before you price.[1] If the geotech is so old, thin, or ambiguous that you don't know the stratum you'd bond into, or the loads and the ground conflict in a way that changes the design, an AI pass that proposes a size against an assumption is not a substitute for resolving the actual question. The tool can flag the RFI — it should — but it can't drill a verification boring or run a proof test. Physical ground truth beats any model.
4. When you already have a faster, battle-tested path for this exact job type
If you've priced this exact niche a hundred times and the set is a template you can bang out from muscle memory with your own rate database, the markup on your standard workflow is probably better than anything a tool adds. Bid intelligence earns its keep on volume and novelty — screening many sets, surfacing gaps in sets you haven't seen — not on repeating a job you already know cold in half the time.
5. When the stakes are "make it up"
If the job is so under-specified that the only honest answer is "we don't know," no tool should paper over that with a confident number. A good bid-intelligence tool flags the ungrounded spots and drafts the RFI. It should never hand you a plausible-looking estimate for a set that can't support one. If a tool gives you a clean number on a set that's a guess, that's the flag to walk away — with or without the tool in the stack.
What this pass list is doing for you
A few things worth being explicit about, because they're easy to miss:
- This is not a disclaimer to bury in the fine print. It's the operating frame. A tool that knows its edges is a tool you can hand work to, because you're not inheriting a black box that might quietly overreach.
- "Pass" is a feature of the same screen. The discipline that says "this job is a no-go" and "this isn't a fit" is the same discipline that should say "this job doesn't need me." Consistency keeps the whole pipeline honest, not just the wins.
- The tool's job is the assist, not the authority. Design output goes to a licensed PE; quantities stay with the estimator's own takeoff and rate database; physical truth and the go/no-go call stay with you.
The test: if it says "use me for everything," walk away
As a rule of thumb: any AI estimating product that presents itself as the right tool for every bid — that never says "this one isn't a fit," never flags what it can't ground, and never mentions engineer review — is not a tool you can rely on. The category earns trust by setting boundaries. The tool you want is the one that tells you honestly where it helps and where it doesn't, the same way a good estimator screens bids.
Want an honest read on your work?
The fastest way to judge is to run one of your real bid packages through it — and let you, not the tool, decide whether it fits. Book a 30-minute demo and we'll run a real set through a cited analysis, and we'll tell you plainly where the tool genuinely helps and where your standard workflow stays the better answer. The honest pass applies to the demo too: if your typical work is the "measure it directly" case, we'll say so and you've lost nothing but a half-hour.
Sources
[1] CMAA — "Navigating Construction Risks With the Go-No-Go Process": understand-the-risk gate before pricing.