The 30-Day
RFP
Thirty days sounds comfortable. It is not. The working time is closer to twenty-one days after weekends, and most of the work depends on people who do not report to you and have other jobs. Teams do not miss RFP deadlines because they are slow — they miss because the schedule was never real.
What follows is a plan that survives contact with reality, followed by the four places it usually breaks and what to do when you are already behind.
The shape to keep
Roughly a third of the calendar for setup and strategy, a third for drafting, and a third for review and finalisation. Teams that lose spend two-thirds drafting and proofread at the end. The back third is where scores are made.
The plan
Qualify and set up
Decide whether you are bidding, and make the document machine-readable.
- Run the bid/no-bid decision before anyone starts writing. Day 1, not day 6.
- Extract every question, requirement, and mandatory into a single tracked list — including the ones buried in prose, appendices, and the terms document.
- Build the compliance matrix: requirement, where it will be answered, who owns it, current status.
- Log every submission mechanic: portal, file formats, page limits, font rules, naming conventions, signature requirements.
- Diarise the clarification-question deadline. It is almost always much earlier than the submission deadline.
Teams routinely spend three days deciding whether to bid, then discover on day 6 that the question list is 40% longer than the visible numbered questions suggested.
Strategy and clarifications
Agree what you are arguing before anyone writes a word.
- Hold a strategy session: what are the two or three reasons this buyer should choose you, and which evaluation criteria do they map to?
- Read the evaluation criteria and weightings closely, then allocate writing effort in proportion to them.
- Submit clarification questions. Ask about ambiguous requirements, page limits, and format constraints — never about anything that reveals a weakness.
- Assign every question an owner and a due date. Owners are individuals, never teams.
Skipping the strategy session is the most common false economy. Without it, twelve people write twelve unrelated arguments and the review cycle becomes a rewrite cycle.
Draft
Get a complete first draft — every question answered, none perfect.
- Start from approved, reused content wherever it exists. Writing from a blank page is a choice, not a requirement.
- Answer in the RFP’s own order and numbering, using the buyer’s terminology.
- Flag every unresolved evidence gap inline rather than writing around it — a visible gap gets filled, a smoothed-over one ships.
- Book SME time now, in calendars, for the specific questions they own. SME availability is your binding constraint.
- Hold a short daily check on the compliance matrix: how many questions are unstarted, drafted, reviewed?
Aim for complete over polished. A draft that is 100% present and 70% good can be improved by review. A draft that is 70% present and 100% good cannot be finished in the time left.
Review
Two distinct reviews with different jobs, not one blended one.
- Compliance review: does every requirement have an explicit, findable response? Run this against the matrix, not by reading.
- Quality review: is every claim evidenced, is the strategy visible, does each answer address exactly what was asked?
- Use different reviewers for the two passes. A compliance reviewer reading for quality misses both.
- Give reviewers a scoring sheet mirroring the buyer’s criteria, and have them score you as an evaluator would.
- Rewrite against review findings. Reserve at least two full days for this — reviews that produce no rework were not reviews.
This is where the schedule is stolen from. Every day the draft slips comes straight out of review, which is the phase that adds the most points per hour spent.
Finalise
Reconcile, price, and approve.
- Cross-volume reconciliation pass, with a single named owner: team sizes, names, dates, durations, totals.
- Lock pricing and confirm it matches every commitment made in the technical volume.
- Executive approval and any legal sign-off. Book this slot in week one — executive calendars do not open at short notice.
- Produce the final document in the required formats and check page limits, fonts, and file naming against the rules you logged on day 3.
Pricing and executive approval are serial, not parallel. An executive who wants a pricing change on day 25 costs you the entire buffer.
Submit
Submit early enough that a failure is recoverable.
- Submit a complete draft to the portal on day 28. Most portals allow replacement until the deadline.
- Verify the upload: open every file from the portal, not from your drive, and confirm it renders and is complete.
- Confirm every required attachment, signature, and form is attached and current.
- Keep days 29 and 30 for final polish and the replacement upload.
Portals fail, files corrupt, and formats get rejected. Submitting on the final afternoon means a technical failure is fatal rather than annoying.
Where it breaks
The plan above is not difficult. It fails in four predictable places, and knowing them in advance is most of the defence.
The question list is longer than you think
Numbered questions are the visible part. Requirements hide in prose ("the supplier shall..."), in appendices, in the terms and conditions, and in the pricing schedule. Teams that count only numbered questions routinely find 30–50% more requirements in week two — after the schedule was built.
Fix: Extract completely on day 1–3 and count before you plan. A wrong denominator makes every subsequent estimate wrong.
SME availability is the real constraint
Your solution architect has a day job and three other bids. The security lead is mid-audit. Draft phase does not fail because writing is slow; it fails waiting on eight people who each owe you two hours.
Fix: Book named calendar slots in week one for specific questions. Give SMEs a draft to correct rather than a blank page — reviewing is roughly four times faster than composing.
Review gets compressed to nothing
Review is scheduled last and squeezed first. A five-day review becomes two days becomes a proofread. Since review is where the largest score gains per hour live, this is the most expensive compression available.
Fix: Treat the review start date as immovable. If the draft is not done, review what exists on the scheduled date and let the gaps be visible.
Nobody owns cross-volume consistency
Technical proposes four people, pricing funds three. Methodology says two-week sprints, the plan shows monthly milestones. Each section is individually correct and the set is incoherent — and evaluators cross-check exactly this to justify scores.
Fix: Name one person, give them a fixed slot on days 22–23, and give them authority to change either section.
A worked example: where the days actually went
Break points are easier to accept in the abstract than to spot in your own schedule. Here is a real shape: a 22-person software firm responding to a county government RFP for a permit-management system. Thirty calendar days. The team had run this process before and considered themselves organised.
| Phase | Planned | Actual | What happened |
|---|---|---|---|
| Qualify & set up | Days 1–3 | Days 1–6 | Bid decision took until day 4; extraction finished day 6 |
| Strategy & clarifications | Days 4–6 | Days 7–8 | Compressed to two days; clarification deadline was day 8 |
| Draft | Days 7–16 | Days 9–21 | Question count grew from 84 to 121; two SMEs unavailable week one |
| Review | Days 17–21 | Days 22–24 | Five days became three |
| Finalise | Days 22–26 | Days 25–28 | Exec wanted a pricing change on day 27 |
| Submit | Days 27–30 | Day 30, 4:40pm | Twenty minutes before the deadline |
They submitted. They lost, scoring mid-field on technical and bottom-quartile on management approach. The instinct afterwards was that the deadline was too tight.
It wasn't. Trace the damage backwards and the whole loss is set in motion in the first week.
The three-day slip in setup is the entire story. Extraction finished on day 6 rather than day 3, because the numbered question list said 84 and the real requirement count was 121 — the extra 37 sat in the terms document, the pricing schedule, and prose "the vendor shall" clauses. That is break point one, exactly: the denominator was wrong, so every estimate built on it was wrong.
Everything downstream inherits those three days, and nothing gives them back. Strategy got squeezed to two days because the clarification deadline is a fixed external date that does not move for you. Drafting absorbed both the delay and 44% more questions than planned, against SMEs whose calendars had been booked for the original scope.
Review paid for all of it. Planned five days, got three — and those three ran concurrently with drafting still finishing, so it was really one clean pass. The compliance check happened once, at the end, and found two requirements from the terms document with no explicit answer. There was no time to do more than paste in something adequate.
Now look at where they actually lost points: management approach, bottom quartile. That is the section that gets written last, reviewed least, and reconciled never. It is the predictable casualty of a compressed back third, and it had nothing to do with how well the team writes.
What one change would have fixed
Not more hours, and not an earlier start. Complete extraction by day 2 — finding all 121 requirements before building the schedule rather than discovering 37 of them in week two. The plan would then have been built against the real workload: either more drafting days allocated from the start, or a scope conversation on day 2 while there was still room to have it. Review keeps its five days either way, and the section they lost on gets read by someone.
This is why extraction is the one step that gets faster but never partial. It is not administrative throat-clearing before the real work — it is the measurement that determines whether your schedule is a plan or a wish.
The hard case: you are on day 18 and the draft is half done
This happens, and the instinct — everyone works late until it is finished — is the wrong move. It produces a complete document with no review, which is reliably worse than an incomplete document that was reviewed where it counts.
Triage in this order.
Behind-schedule triage
- 1Every mandatory requirement gets a complete, compliant answer. Non-compliance is disqualification, and a disqualified proposal scores zero regardless of quality.
- 2Every question carrying heavy evaluation weight gets full effort. Check the weightings — a 25%-weighted section deserves five times the attention of a 5% one, and teams routinely spread effort evenly instead.
- 3Low-weight questions get a competent, short, reused answer. Adequate and present beats excellent and absent.
- 4Protect two days of review no matter what. Cut drafting scope to buy them back.
- 5Cut optional material first: extended case studies, additional appendices, design polish. Nobody has ever lost on appendix formatting.
- 6If genuinely unrecoverable, withdraw rather than submit something weak. A visibly poor submission damages you with that buyer for the next cycle; a courteous withdrawal does not.
The judgement call is step six, and it is genuinely hard. The rule of thumb: if you cannot answer every mandatory requirement compliantly, you are not submitting a proposal — you are submitting a rejection notice with your logo on it. Withdraw, and tell the buyer why. Teams that do this are remembered as professional rather than absent.
Shorter deadlines
Two-week windows are increasingly common. The proportions hold — roughly a third, a third, a third — but the setup phase compresses hardest, which is exactly the phase that determines whether the rest of the schedule is real.
On a fourteen-day cycle, the qualification decision has to happen on day 1 and extraction has to be finished by day 2. There is no slack for discovering hidden requirements in week two. This is why short-deadline RFPs are a different bid/no-bid question: a bid you would take at thirty days may be a no-bid at fourteen, purely on response capacity.
What makes the plan survivable
Look at where the days actually go and a pattern appears. Extraction, hunting for content you have already written, chasing status across a spreadsheet, and reconciling the same facts across volumes — none of that is thinking work, and all of it lands squarely on the phases that get compressed.
When extraction is complete and automatic on day one, when approved answers surface instead of being hunted, and when compliance status is visible rather than reconstructed weekly, the back third of the schedule survives. That is the whole game: protect review, because review is where the score is made.
Protect the back third.
Complete extraction on day one, approved content that surfaces itself, and live compliance status — so review never becomes a proofread.
Get Started for Free