The Bid/No-Bid
Decision

Most teams trying to raise their win rate work on their writing. That is usually the wrong lever. The largest single driver of win rate is not how well you answer — it is which RFPs you agree to answer at all.

A team that bids everything and wins 18% is not a team with a writing problem. It is a team spreading a fixed amount of subject-matter-expert attention across too many documents, so every response is thin. The same team bidding half as often, on better-qualified work, routinely lands between 35% and 45% — with less overtime.

The hard part is that "no-bid" feels like giving up, and pipeline pressure makes every RFP look like an opportunity. The fix is to make the decision mechanical, made against written criteria before anyone gets attached, rather than argued case by case in a meeting where the loudest optimist wins.

First, the arithmetic nobody runs

Before the framework, it helps to price the thing you are deciding about. A mid-size response is not free, and the cost is mostly senior time — the scarcest resource you have.

There is a second cost that never appears on a spreadsheet: the opportunity cost of the SME hours. The twelve hours your lead architect spent on a long-shot bid were not available for the pursuit you were genuinely positioned to win. Bad qualification does not just waste effort — it degrades the responses you should have been winning.

The disqualifiers: five questions that end it

Run these before any scoring. They are binary, and any single "yes" is a no-bid unless an executive overrides it in writing with a stated reason. Scoring a deal that fails one of these is how teams talk themselves into bad pursuits — a strong score on five criteria masks the one that was fatal.

The scoring model

If the deal clears all five disqualifiers, score it. Six criteria, weighted, each rated 1 to 5 against written anchors. Anchors matter more than the weights — without them, everyone scores 3 and the model tells you nothing.

Reading the total

Multiply each score by its weight and sum. The result lands on a 100-point scale, and the thresholds are deliberately unkind:

• Below 55 — no-bid. Not "bid if we have time." No-bid. Send a courteous decline and ask to be included next cycle; declining well is a relationship move.
• 55 to 70 — conditional. Bid only if you can name the specific action that raises a low score before the deadline: a partner who closes the solution gap, a reference call that builds relationship, a pricing structure that fixes commercial value. If nobody will own that action, it is a no-bid.
• Above 70 — bid, and resource it properly. These are the pursuits that deserve your best SMEs and a real review cycle.

A worked example

A 140-person managed-services firm receives a three-year infrastructure support RFP from a regional hospital network. Contract value roughly $2.1M. The team wants to bid — it is the largest opportunity in the pipeline this quarter.

Fifty-nine. Conditional — and the shape of the score is more informative than the number. The two heaviest criteria, relationship and competitive position, are the two lowest. The enthusiasm was coming entirely from commercial value, which carries 15% weight and describes the prize rather than the odds of winning it.

That is the specific failure mode this model exists to catch: deal size masquerading as deal quality. A $2.1M contract you will probably lose is worth less than a $400K contract you will probably win, and it costs more to chase.

The conditional path here has one honest question. Can anything move before the deadline?

• Relationship (2). Nineteen days out, with the buyer in a quiet period, relationship cannot be built. This score is fixed.
• Competitive position (2). Displacing a six-year incumbent requires either a documented service failure you can point to or a structural cost advantage. The team has neither. Fixed.
• Response capacity (2). Could be raised by pulling SMEs off the other two bids — which are better-qualified. Raising it would lower the odds on work you were more likely to win.

Nothing moves. This is a no-bid, and the discipline is in saying so while the number still says "conditional." The strategic instinct — get into healthcare — was sound. The conclusion is that this RFP is the wrong door into that vertical. The right move is to spend a fraction of that $18,900 on getting known before the next one, so relationship scores 4 instead of 2 when the re-compete comes around.

The hard case: when you have to bid anyway

Sometimes a low-scoring bid is mandatory. A framework agreement requires participation to stay on the vendor list. A major account expects to see you respond. A board wants evidence of activity in a target segment. Pretending otherwise makes the model something people route around, and a model people route around is worse than no model.

Handle it explicitly. Call it a strategic bid, record the real reason, and — critically — cap the investment. A strategic bid gets reused content, a single reviewer, and a hard hour budget. It does not get your best architect for two weeks. The failure is not bidding a long shot; it is bidding a long shot at full price.

Track these separately in your win-rate reporting. A team with 40% on qualified bids and 5% on strategic bids has a clear, healthy picture. Blending them into one 22% number tells you nothing and hides both signals.

Making the decision stick

Three operational details separate teams that use a framework from teams that have one in a folder.

Score before the kickoff meeting, not during it. Once the team has assembled and started dividing sections, the sunk-cost reflex sets in and no-bid becomes socially expensive. The decision belongs on day one, ideally by two people scoring independently and comparing.

Require evidence, not adjectives. Every score needs a sentence of justification in the evidence column. "Relationship: 4" is an opinion. "Relationship: 4 — we ran a discovery workshop with their ops director in March and reviewed a draft scope" is a fact. Opinions drift upward under pressure; facts do not.

Review the model against outcomes twice a year. Pull every bid from the last six months, plot score against result, and check whether the threshold is calibrated. If you are winning bids that scored 50, your weights are wrong. If you are losing at 80, your anchors are inflated and everyone is scoring generously. The model is a hypothesis about what predicts winning at your company, and it should be tested like one.

What changes when qualification gets tighter

The uncomfortable early effect is that pipeline coverage looks worse. You are submitting fewer proposals, and a metric someone reports on goes down. Get ahead of that by reporting qualified-pipeline value and win rate together from the start, so the trade is visible as a trade rather than a decline.

The effect that follows, usually within two quarters, is that responses get noticeably better — not because anyone learned to write, but because the same SME hours now spread across half as many documents. Depth is a function of attention, and attention is a function of how many things you said yes to.

When you do commit to a pursuit, the mechanical work of a response — extracting every question, finding what you have already answered, keeping compliance visible — should take as little of that attention as possible. That is the part worth automating, so the hours you protected by saying no go into strategy and proof rather than copy-paste.