Every estimating department has a version of this conversation: three weeks into a pursuit, someone finally asks the question that should have killed it on day one — the owner's last project ended in litigation, or the LDs are uncapped, or the peak lands exactly on top of the courthouse job. The information was available the whole time. Nobody asked in sequence.
That is what a checklist is for. Not to make the decision — that is the weighted scorecard's job — but to make sure every killable pursuit dies in the first ninety seconds, before it consumes a meeting, let alone a takeoff. Here are the twenty-five questions, grouped in the order they should be asked, with the automatic kills marked.
Category 1 — Owner & money (ask these first)
- Have we worked for this owner or GC before, and how many days did they actually take to pay? Contract terms say net-30; your AR ledger says the truth.
- Is the project funded? Public: appropriation passed, not proposed. Private: construction loan closed or owner balance-sheet financed. "Funding in progress" is a no-bid until it isn't.
- Any liens, bond claims, or payment litigation on the owner or GC in the last three years? A state lien search takes ten minutes. Unresolved payment litigation = automatic kill.
- What would our peak accounts receivable be, and what fraction of working capital is that? Model it — billing minus lagged receipts minus retention. Above 40% of working capital = automatic kill, per the hard-kill rule in our scorecard tool.
- Who pays for the float? If retention is 10% released at final acceptance, someone finances that money for months after your last labor spend. If it isn't priced, it's donated.
Category 2 — Schedule & capacity
- Does this project's peak manpower land on top of an existing peak? The question is never "can we do it" — it's "what do we displace to do it."
- Are the PM and foreman we would actually assign clean by the start date? Named people, not org-chart theory.
- Do we have bonding capacity left after current backlog? Your surety's answer, not your controller's guess.
- Can we meet the bid date with a real quantity survey? If the only way to make the date is a plug-number bid, that's an automatic kill — a bad number is worse than no number.
Category 3 — Scope & documents
- What design stage are the documents, honestly? "100% CDs" with uncoordinated MEP backgrounds are 80% CDs. Design maturity is the best single predictor of change-order exposure.
- Are the specs project-specific or copy-pasted? Boilerplate specs from another job type signal a design team that will answer RFIs slowly and defensively.
- How many addenda have already been issued? Three-plus addenda before bid day means the design is still moving — and will keep moving after award.
- Is our scope clearly separable? Fuzzy trade boundaries (who owns blocking? roof curbs? fire caulk?) become scope-gap losses at buyout.
- Are there named products with no "or equal"? Sole-sourced specs transfer supplier pricing power directly onto your bid.
Category 4 — Contract terms
- Retention percentage and — more important — the release trigger? 5% at substantial completion and 10% held through final acceptance are different jobs.
- Liquidated damages: amount, cap, and does our float own them? Uncapped LDs on a schedule you don't control = automatic kill for most subs.
- Pay-if-paid or pay-when-paid? Pay-if-paid shifts the owner's insolvency risk onto you; some states void it, many don't.
- Change-order pricing mechanism? If T&M rates and markup percentages aren't fixed in the contract, every change becomes a negotiation you start from zero.
- Termination-for-convenience terms? You want demobilization plus reasonable overhead recovery, not "paid for work in place" only.
Category 5 — Competition & strategy
- How many bidders, and who are they? Three to five qualified names is a market; twelve on an open list is a lottery. See the hit-ratio math on bidder count.
- Is there an incumbent or a designed-around favorite? If the spec reads like a competitor's product sheet, you're the coverage bid. Decline fast with a no-bid letter.
- Does winning this build anything? Repeat-owner potential, a new market foothold, a crew kept whole through a valley — or is it just revenue?
Category 6 — Execution risk
- Site logistics we can't control: occupied facility, hospital infection control, night work, union jurisdiction lines, laydown at a premium?
- Systems or assemblies we've never built? First-time learning curves belong on negotiated work, not hard bids.
- Distance from our operating base? Past the drive-home radius, add per-diem, turnover, and supervision dilution — or don't bid it.
Questions 3, 4, 9, and 16 are pass/fail. If any fails, the pursuit dies at the checklist — no meeting, no scorecard, no debate. A written kill reason goes in the pursuit log and the no-bid letter goes out within 48 hours. Everything else is input for the scorecard, not a verdict by itself.
From checklist to score
A pursuit that survives the screen graduates to the seven-question weighted scorecard — pay history, schedule fit, and AR exposure at 3x weight; personnel, scope clarity, and retention at 2x; competitive field at 1x. That's where "mostly fine with two concerns" becomes an actual number against calibrated thresholds. Run it in the interactive scorecard or the Excel workbook — both apply the hard-kill rules automatically and both take about five minutes.
The checklist-then-scorecard sequence has one more benefit nobody advertises: it makes the go/no-go meeting short. When the screening questions are answered before the meeting and the scorecard is filled in during it, the whole decision takes fifteen minutes — which is the only speed at which a precon desk will actually keep doing it.
PILARS exists for the pursuits that pass: AI takeoff at $100 per trade per plan, so the estimating hours you protected go into strategy instead of counting. See pricing here.