The short answer
A bid/no-bid decision framework is a scoring sheet plus a deadline. Score every new RFP on seven weighted criteria (buyer contact before release, fit to mandatory requirements, competitive position, capacity, margin, strategic value, contract risk) for a total out of 100. Bid at 70 or above, escalate 50 to 69, decline below 50. A short list of kill gates overrides the score, and one named decider calls it within two business days of release.
Key takeaways
- Qualification is common but slipping: Loopio’s 2026 benchmark survey found 75% of teams use a go/no-go process, 8 points fewer than in 2024.
- Score, then gate: A weighted score ranks opportunities. Kill gates (a mandatory requirement you can’t meet, uninsurable liability, no team) stop a bid no matter how good the score looks.
- Decide fast: Call it within two business days of release. Every day of dithering comes out of the writing window, not the buyer’s.
- One decider, many inputs: Sales, delivery, finance and legal each score their own rows. One person signs the decision and owns it.
- Track the no-bids: Reason codes on every declined RFP tell you which markets, buyers and contract terms you should stop chasing.
- The catch: The thresholds here are a starting rule, not a law of nature. Until you have 15 to 20 decided bids with outcomes, treat the bands as a conversation starter.
Why qualification beats writing faster
Every RFP you answer costs real hours. Loopio’s 2026 report puts the average at 33 hours per RFP, and says organizations respond to 55% of the RFPs they receive, down from 63% the year before (Loopio, March 26, 2026). The same survey says 81% of its top performers (teams winning more than half their submissions) use a go/no-go process, against the 75% average.
That’s a vendor-run survey, so read it as a directional signal rather than a controlled study. The arithmetic doesn’t need a survey, though. If your team has capacity for 10 responses a quarter and three of them go to RFPs that were wired for an incumbent, you’ve spent 30 percent of your quarter on paperwork that was never going to win. Dropping those three doesn’t just save hours. It moves your best writers onto the bids you can win.
The other quiet benefit is honesty with sales. A written score turns “I’ve got a good feeling about this one” into a row that says “no contact with the buyer before release: 1 out of 5.” Feelings are hard to argue with. Rows aren’t.
The scoring sheet
Copy this into a spreadsheet tab called Bid_Decision and keep one row per criterion. Score each criterion from 0 to 5, then multiply by the weight and divide by 5. The weights below total 100, so the result reads as a score out of 100.
| Criterion | Weight | 0 looks like | 5 looks like | Who scores it |
|---|---|---|---|---|
| Buyer contact before release | 20 | First you heard of it was the portal notice | You shaped requirements in discovery or answered the RFI | Account owner |
| Fit to mandatory requirements | 20 | One or more pass/fail requirements you can’t meet | Every mandatory item met with current product and evidence | Bid manager plus solutions lead |
| Competitive position | 15 | Happy incumbent, requirements read like their datasheet | No incumbent, or an incumbent the buyer is openly replacing | Account owner |
| Capacity in the response window | 15 | Writers and SMEs already booked on committed bids | Named writer, SMEs and reviewer free for the full window | Bid manager |
| Margin | 10 | Price to win sits below your floor | Price to win sits well above your floor | Finance or deal desk |
| Strategic value | 10 | One-off deal in a market you’re leaving | Logo, segment or reference you’ve named as a goal | Sales leader |
| Contract risk | 10 | Unlimited liability, punitive SLAs, IP terms legal rejects | Standard terms, or terms you’ve accepted before | Legal |
A few scoring rules keep the sheet honest:
- Score with evidence, not optimism. Each score cell gets a comment: a meeting date, a requirement ID, a margin calculation. No evidence, maximum score of 2.
- Score fit against the mandatory list, not the whole RFP. Pull the “shall” and “must” statements into a list first. The free RFP requirement extractor does this in a minute, and the compliance matrix guide covers what to do with the list afterward.
- Let the owner score their own row. Legal scores contract risk, finance scores margin. The bid manager doesn’t guess on their behalf.
- Write the total and the band on the first row. Go (70 to 100), conditional (50 to 69), no-bid (below 50).
Why weight buyer contact so heavily? Because an RFP is often the last step of a buying process, not the first. If the buyer never spoke to you before release, the requirements were shaped by somebody else’s conversations. You can still win from a standing start. It’s just rarer, and the sheet should say so.
Kill gates that override the score
A high score can hide a fatal flaw. Kill gates are yes/no questions checked before the total counts. Answer “yes” to any of them and the answer is no-bid, unless the named decider overrides it in writing with a reason.
- Mandatory miss: You fail a pass/fail requirement (a certification, a hosting region, a licence, a minimum years-in-business rule) and can’t fix it before the deadline.
- Eligibility: You aren’t eligible to bid at all, for example a set-aside or framework you don’t hold.
- No team: You can’t name a writer, a solutions lead and a reviewer for the window without pulling them off a bid you’ve already committed to.
- Unacceptable terms: Legal won’t sign the contract terms as written and the RFP says exceptions will be marked non-compliant.
- Below the floor: Your best estimate of the price to win sits below your walk-away margin.
- Conflict: A conflict of interest, an existing dispute with the buyer, or a reference you can’t provide.
Keep the list short. Six gates you actually enforce beat twenty nobody reads. The BAFO guide explains how to set the walk-away floor this gate depends on.
Who decides, and by when
The framework fails most often on timing, not scoring. A decision on day nine of a 20-day window leaves the writers eleven days for a 20-day job.
| Step | Owner | Deadline (from release) |
|---|---|---|
| Log the RFP, extract mandatory requirements, set the decision meeting | Bid manager | Same business day |
| Score own rows with evidence | Account owner, solutions lead, finance, legal | End of business day 1 |
| Decision meeting (15 minutes, no slides) | Named decider, usually the sales leader for the segment | Business day 2 |
| Kick-off or decline note | Bid manager | Business day 2 or 3 |
Two business days works for windows of roughly three to six weeks. For a window under two weeks, run the whole thing in one day. For a large federal or multi-volume bid with a 60-day window, you’ll often have qualified the opportunity months earlier during capture, so the release-day decision is a confirmation against the final text.
The decider has three options, not two: go, no-go, or go with conditions. Conditions must be specific and dated. “Go, provided the solutions lead confirms the data-residency requirement by Thursday” is a condition. “Go, but be careful” isn’t.
One more rule worth stealing: a go decision can be reversed. If a clarification answer or an amendment changes a kill gate, rerun the sheet. Pulling out on day eight is cheaper than submitting a non-compliant response on day 20. The clarification questions guide covers how to ask the question that settles it.
What US federal rules say about early contact
If you sell to US federal agencies, the regulation explicitly expects you to make this call early. FAR 15.201(b) says the purpose of exchanges with industry before proposals are received is to improve understanding of requirements, “thereby allowing potential offerors to judge whether or how they can satisfy the Government’s requirements” (FAR 15.201, FAC 2026-01). Paragraph (c) encourages agencies to “promote early exchanges of information about future acquisitions.”
Some agencies go further and tell you directly. Under the advisory multi-step process in FAR 15.202, the agency evaluates short submissions and advises each respondent in writing that it will be invited to participate or that it is “unlikely to be a viable competitor,” while also informing respondents they may still participate. That letter is a gift for your scoring sheet. Treat an “unlikely” verdict as a 0 on competitive position unless you have strong evidence the agency is wrong.
A note on versions: the FAR is mid-rewrite under the federal FAR overhaul (RFO). The citations above are the codified FAR at FAC 2026-01, effective March 13, 2026. The Part 15 model deviation text issued September 30, 2025 moves the same ideas to 15.101 (early exchanges, including the advisory multistep process at 15.101(e)), and agencies that adopt it by class deviation will cite those numbers instead. The government RFP guide covers the rest of the federal process.
Worked example: Scoring an illustrative RFP
Say you sell a mid-market claims platform. An illustrative regional insurer releases an RFP on a Monday with a 25-business-day window. Here’s how the sheet might fill in by Tuesday afternoon.
| Criterion | Weight | Score (0 to 5) | Weighted (weight x score / 5) | Evidence noted |
|---|---|---|---|---|
| Buyer contact before release | 20 | 2 | 8 | One discovery call 5 months ago, no follow-up |
| Fit to mandatory requirements | 20 | 4 | 16 | 41 of 43 mandatory items met; 2 need a roadmap answer |
| Competitive position | 15 | 2 | 6 | Incumbent in year 6 of a 5-year contract plus extension |
| Capacity in the response window | 15 | 3 | 9 | Writer free; security SME shared with another bid for week 2 |
| Margin | 10 | 4 | 8 | Estimated price to win clears the floor by a comfortable gap |
| Strategic value | 10 | 5 | 10 | Insurance is a named target segment this year |
| Contract risk | 10 | 3 | 6 | Liability cap at 2x fees; legal can live with it |
| Total | 100 | 63 | Conditional band |
Kill gates: the two unmet mandatory items are the only candidates. The RFP marks one of them as “desirable,” not mandatory, on a second read. The other (a single-sign-on protocol) is on the roadmap for next quarter, which the RFP’s go-live date would accommodate. No gate fails.
At 63, the sheet lands in the conditional band, so the decider sets two dated conditions:
- The account owner gets a call with the buyer’s project lead by Thursday, or submits a clarification question about the single-sign-on timing through the portal by the question deadline.
- The bid manager confirms the security SME can cover week 2 by moving a lower-scored DDQ deadline.
Run the arithmetic both ways. If the call happens and the buyer confirms they’re replacing the incumbent, competitive position goes from 2 to 4 (+6 points) and buyer contact from 2 to 3 (+4 points), taking the total to 73: a clear go. If the account owner can’t get the call and the SME conflict stands, capacity drops to 2 (minus 3 points) and the total falls to 60. Still conditional, but the decider now has a reason on paper to decline, and nobody has written a page yet.
That’s the point of the exercise. The number isn’t the decision. It’s the shared, written basis for one.
The decline note
Send it to the procurement contact named in the RFP, before the deadline, through the channel the RFP specifies for correspondence. Keep it under 150 words. A template:
Subject: [RFP reference]: decision not to submit
Hi [name],
Thank you for inviting [company] to respond to [RFP title]. After reviewing the requirements and timeline, we’ve decided not to submit a proposal this time.
[One honest, non-critical reason, for example: “The required go-live date falls before our next release of the single-sign-on capability you’ve listed as mandatory.”]
We’d welcome the chance to be considered for future work, and we’re happy to share feedback on the requirements if that would help your team.
Best regards, [name, title, direct contact]
Don’t criticize the RFP, the timeline or the incumbent. Don’t promise to bid next time. And don’t send it if the RFP forbids contact outside a portal; post it there instead.
Tracking no-bid reasons
A no-bid you don’t record teaches you nothing. Add these columns to the same log you use for submitted bids, so the win-rate metrics and the no-bid patterns live side by side.
| Column | Example value |
|---|---|
rfp_id |
2026-INS-014 |
release_date / decision_date |
October 5, 2026 / October 7, 2026 |
total_score |
47 |
failed_gate |
Below the floor |
primary_reason_code |
PRICE, FIT, CAPACITY, INCUMBENT, TERMS, ELIGIBILITY, LATE_NOTICE |
decline_note_sent |
Yes |
follow_up |
Ask for award notice and winning price where published |
Review the log every quarter. Three questions are worth asking:
- Which reason code dominates? If CAPACITY tops the list, you have a hiring or tooling problem, not a market problem. The deadline bottlenecks guide is the next read.
- Did any no-bid buyers award to someone you beat elsewhere? That’s a sign the gate or the weight was too strict.
- Did your go decisions in the conditional band win? If they lose far more often than the 70-plus bids, raise the go threshold.
Where software helps (and where it doesn’t)
Software is good at the reading. It’s weak at the deciding.
The jobs it does well are the mechanical first pass: pulling mandatory requirements out of a 120-page package, matching them against your documentation, and flagging the gaps before the decision meeting. Tendrio is built for that front end: it turns a tender package into a compliance matrix marked met, partially met or gap, and offers go/no-go recommendations from that matrix. Its own scope note is worth repeating, though: there’s no answer library or approval workflow, and sign-off happens outside the tool. Tenderbolt runs a go/no-go analysis of tender documents with scored, source-linked reasoning, and suits European teams that want data kept in the EU. Iris offers go/no-go scoring for RFP fit against win history, aimed at presales teams.
For US federal work, capture is a bigger job than a single go/no-go call. GovDash and Sweetspot both carry the capture-management tag: GovDash runs one workflow from opportunity through proposal, and Sweetspot pairs an opportunity engine over federal award data with RFP shredding into a compliance matrix. Tendium covers the European public-sector version, with tender monitoring and Smart Checklists for the bid/no-bid read.
What none of them can do is know that the buyer’s new CIO used to be your customer, that your best SME is about to go on leave, or that your CFO will accept a thinner margin to land the segment. Those inputs carry most of the weight in the sheet above, and they live in people’s heads.
If you see fewer than five RFPs a month, skip the tooling for this step. The spreadsheet tab above, a shared calendar and a 15-minute meeting will do the job. Buy software for the reading when the reading, not the deciding, is what’s slowing you down.
Sources
- Average RFP Win Rates & More: 38 Proposal Statistics to Know in 2026, Loopio, March 26, 2026, accessed October 2026
- 15.201 Exchanges with industry before receipt of proposals, Federal Acquisition Regulation (FAC 2026-01, effective March 13, 2026), acquisition.gov, accessed October 2026
- 15.202 Advisory multi-step process, Federal Acquisition Regulation (FAC 2026-01), acquisition.gov, accessed October 2026
- FAR Overhaul: Part 15, RFO model deviation text (issued September 30, 2025), acquisition.gov, accessed October 2026