Federal Proposal Capture: What Small Teams Do Before the RFP Drops

Tuesday at 4:30: what would you do?

A SAM.gov alert lands. Partner email from this morning calls the same notice a “great capture.” Contract value looks healthy. You have ten minutes between calls. There is no capture shop, no color-team calendar, and no one else to own the next hour of B&P.

You have not decided whether this pursuit deserves any more of your week.

What would you do?

The correction: Option C wins. Option A fails because process without leverage is expensive cosplay for a 15-person shop. Option B is the most common small-team trap: excitement later, showstoppers later, sunk hours third. For Sam, “winning” starts as better shots and fewer doomed bids.

The Pursuit Napkin (what good early capture looks like)

Experienced small teams do not need a 40-factor scorecard. They need a short, versioned napkin that answers five questions before Write starts: Can we even do this? What public facts change our odds? Who must we team with? Is an RFI worth the hours? Are we Go, Watch, or No-Bid until paper drops?

That is federal proposal capture when you wear every hat. The industry habits still matter (qualify hard, gather citeable intel, team to fill gaps, engage public market research when leverage exists, freeze a decision). You just skip the binder.

Pain you avoid when you do this first: partner-driven “must bid” pressure with no receipts, RFI responses that invent intimacy you cannot prove, and Friday discovery that clearance, geography, or past performance rules already killed the shot.

Habit 1: Qualify before you crush on

Write three lines: capability fit, people available in the response window, and hard realities (clearance, place of performance, certifications you hold at submission). If any line is a hard no, stop. That is capture doing its job.

Habit 2: Public intel only

Pull what is citeable: SAM history, USASpending or FPDS awards, incumbent name if public, agency strategy docs. If you cannot point to a source, it does not go on the napkin. Fantasy intel is how small teams talk themselves into Write.

Habit 3: Teaming as gap-fill, not logos

List the gap you cannot fill alone, what a partner must bring, and your walk-away if the partner wants prime. Logos without a gap are vanity. A single missing Key Personnel or past performance lane is often the real gate.

Habit 4: RFI / Sources Sought only when it buys leverage

Public market research is worth hours when you can clarify requirements, show relevant capability, or learn evaluation direction. It is not worth hours when the notice is a checkbox, you have nothing distinct to say, or you would be inventing “insights” the agency did not ask for. Stay on public channels only.

Habit 5: Bid/no-bid freeze

Freeze Go / Watch / No-Bid with a reason Sam can defend to a partner in one sentence. Set a review date (for example, draft RFP drop or RFI due date). Without a freeze, every alert stays “maybe” forever and B&P leaks.

Habit 6: When paper drops, hand off to Decide depth

An RFI package, draft RFP, or final RFP is a Decide event. Run Outside-In hard gates before the SOW. Then size effort with cost-to-bid / Section L and M workload math. If influence risk is the worry, use the wired RFP red-flag screen. Do not re-teach those essays here; link and move.

Filled Pursuit Napkin (sample)

One realistic vignette for a fake agency IT services recompete (your napkin stays this short):

AI to the rescue: techniques and prompt starters

General-purpose LLMs are useful as faster interns for napkin drafts and RFI skeletons. They do not replace SAM dates or the PDF. They do not create customer relationships you do not have.

Techniques that hold up:

• Public sources only: paste notice text, RFI questions, or award excerpts. Ban “assume we know the PM.”
• Citation discipline: every napkin claim needs a source line. If the model cannot cite, mark NEEDS_REVIEW.
• Draft is not final: treat draft RFP language as provisional; re-run gates when the final package lands.
• Skeptic pass: ask what would make this a No-Bid for a 15-person shop.

Prompt starter 1: Pursuit Napkin from a SAM notice

Prompt starter 2: public intel checklist

Prompt starter 3: RFI reply skeleton

Prompt starter 4: bid/no-bid freeze helper

Common AI failure modes (early capture): invented customer intimacy, fake incumbents, treating draft RFP as final, and confident RFI essays that overclaim. Fix with narrower pastes, forced cites, and human verification on SAM and PDFs.

Verification (non-negotiable)

Before anyone treats the napkin as “done,” a human confirms on SAM.gov (notice type, response date, timezone, active status) and on the source PDFs for any RFI, draft, or final package. Public solicitations only. Do not upload CUI, ITAR, or Classified. AI proposes; Sam owns Go/No-Go.

Why even great napkins are not the whole system

The workflow above works until a draft RFP ZIP lands with five volumes, two amendments, and a partner pushing kickoff the same afternoon. Consistency breaks under package size, amendment churn, and handoffs. That is not a skill gap. It is an orchestration gap on the Decide step.

Next in this series

Early capture freezes the shot. Hard gates and cost-to-bid LOE decide whether Write is worth staffing. For finding the next alert, use SAM.gov saved search setup.

Do not skip verification because the napkin “looks decisive.” The expensive failures are almost always one amendment or one mandatory deep.

Frequently asked questions