How to Measure and Improve Your RFP Win Rate

Win rate is the number every proposal team quotes and almost no proposal team defines the same way. Two firms can both claim “a 40% win rate” and mean entirely different things – one is measuring submitted proposals that were awarded, the other is measuring every opportunity that entered the pipeline. Until you fix the definition, you cannot improve the number, because you are not measuring the same thing twice.

This guide is the hub for measuring and improving RFP win rate. It covers how to define and calculate the right numbers, which leading indicators predict the outcome, how to run a win-loss program that tells the truth, what a proposal actually costs, what good looks like by segment, and the pipeline decisions that move win rate faster than any rewrite.

Win rate is an output. The inputs you control are selectivity, compliance, evidence and learning. Improve the inputs and the output follows.

Why your win rate is probably misread

Most confusion comes from the denominator. There are at least three legitimate “win rates,” and they answer different questions:

• RFP win rate – proposals won divided by proposals submitted. This measures execution quality on the bids you chose to pursue.
• Sales win rate – deals won divided by all opportunities, including the ones you never bid. This is a much lower number because it includes opportunities you filtered out. Average B2B sales win rates commonly sit around 20%, while RFP win rates commonly sit around 45% (Loopio, 2026). They are not contradictory; they measure different stages.
• Competitive win rate – proposals won divided by competitive decisions (excluding no-decisions). This isolates head-to-head performance.

Quoting one number without naming the denominator is how teams end up arguing past each other. Decide which number you mean, define it in writing, and measure it consistently. For the calculation methods, see How to Calculate Proposal Win Rate.

The four numbers that matter

A single win rate hides where you are strong and where you are weak. Track a small dashboard instead.

Read them together. A high win rate with a low pursuit rate may simply mean you bid rarely and pick well. A high pursuit rate with a low shortlist rate means you are responding to too many bids you cannot win. And a competitive win rate that trails your overall win rate suggests you are losing the deals where a real competitor shows up. For the shortlist distinction, see Shortlist Rate vs Win Rate.

The value of the dashboard is diagnosis. When win rate dips, the other three tell you where to look. A falling shortlist rate points to the written response. A falling pursuit rate points to pipeline coverage drying up. A widening gap between overall and competitive win rate points to positioning against a specific competitor. One number tells you something changed; four numbers tell you what to do about it.

Interpretation also depends on your bid mix. A firm that mostly renews incumbent contracts should expect a materially higher win rate than one competing for cold, competitive bids – even with identical execution quality. That is why segmenting the numbers matters more than optimizing the blended figure, and why a falling headline number is not automatically bad news if the mix has shifted toward harder, higher-value work.

How to calculate win rate (with a worked example)

The blended formula is simple:

Win rate = (proposals won / proposals submitted) x 100

If you submitted 60 proposals and won 24, your win rate is 40%.

Three refinements make the number honest:

• Exclude undecided bids. Leave out anything still open or cancelled; only count decided outcomes.
• Measure by segment. Report win rate by deal size, industry, incumbent versus new business, and competitor. A single blended number hides the pattern that matters.
• Weight by revenue where it counts. A revenue-weighted view reveals whether you win small deals but lose the large ones, which a headcount-based percentage obscures.

The most common mistake is averaging across incompatible segments. An incumbent renewal that wins at 60-90% is a different animal from a cold, competitive bid that wins at single digits to mid-teens. Blending them produces a number that drives the wrong decisions.

Leading indicators: shortlist rate and evaluator score

Win rate is a lagging indicator. By the time you know it, the bids are long decided. To act while it matters, track what predicts the outcome.

• Shortlist rate shows whether your written response survives the first cut. AutoRFP.ai’s 2026 report found high-win teams achieve a median shortlist rate of 63%. If your shortlist rate is low, the problem is in the response, not the pricing conversation.
• Evaluator score and feedback shows how close you were. Where buyers provide scores or debrief notes, capture them; a pattern of “compliant but unremarkable” points to a differentiation problem, not a compliance one.
• Requirement coverage shows whether you are losing on technicalities. A rising count of missed mandatory requirements is a process failure you can fix this week.

Track these monthly. They move before win rate does, which is what makes them useful.

Run a win-loss program that tells the truth

Most teams think they know why they lose. The evidence says they usually do not. Clozd’s win-loss research found that CRM-recorded loss reasons disagree with what buyers actually say more than 60% of the time, and that a different competitor was identified in roughly seven of ten deals. Corporate Visions, analyzing over 100,000 B2B purchase decisions, found that sellers and buyers cite different reasons 50-70% of the time – and that 53% of deals marked “lost” were winnable but for a fixable misstep in the sales process.

The implication is uncomfortable but useful: if you are improving based on rep-reported loss reasons, you are probably fixing the wrong things. A structured win-loss program replaces guesswork with buyer evidence.

How to run one that produces signal rather than anecdote:

• Interview buyers, not just reps, through a neutral party – not the account executive who lost the deal.
• Sample deliberately: a balanced mix of wins, losses and no-decisions, weighted toward competitive and strategic deals.
• Reach a useful sample size. Roughly 10-15 interviews per quarter is a practical floor for a lean team; fewer than eight produces anecdotes stakeholders can dismiss.
• Interview within two to four weeks of the decision, while recall is fresh.
• Code and count. Convert interviews into themes, then rank fixes by frequency, revenue impact and how addressable they are.
• Act and re-measure. The point is not the report; it is the change.

The payoff is documented: Clozd’s State of Win-Loss analysis found 84% of programs running two years or more reported a win-rate increase. For the full playbook, see Win-Loss Analysis: The B2B Services Playbook.

Cost per proposal and proposal ROI

You cannot judge a win rate without knowing what each response costs. Benchmarks put a single response at roughly 25-41 hours of work, and loaded labor cost per proposal commonly falls between $2,000 and $10,000, with mid-market teams often higher once subject-matter expert time, reviews and overhead are included (APMP and Loopio benchmark data). Complex, compliance-heavy bids can exceed that comfortably.

Turn cost into a decision by computing three numbers:

• Cost per proposal – fully loaded labor and overhead per response.
• Cost per win – cost per proposal divided by your win rate.
• Proposal ROI – the value of won work per dollar of proposal cost.

These reframe the debate. A team with a 25% win rate and a $6,000 cost per proposal spends $24,000 in proposal cost for every win. Raising the win rate to 33% cuts that to $18,000 without changing the per-proposal cost – which is why selectivity and quality usually beat simply bidding more. For the full method, see Cost per Proposal and Proposal ROI.

Benchmarks: what good looks like

Benchmarks are directional, not destiny, but they give you a reference point. The clearest published figures:

Two cautions. First, the numbers move with the mix: a firm weighted toward incumbency will outperform one weighted toward cold bids, even with identical execution quality. Second, a rising market average does not mean your number should rise; measure against your own prior quarters and against your segments, not against a headline. For the full benchmark discussion, see What Does a Good Win Rate Look Like?.

The no-decision problem

A large share of “losses” are not losses to a competitor at all. They are deals that simply stop, because the buyer decides not to change anything. Research on millions of B2B sales conversations puts the share of deals ending in no decision at 40-60%. In proposal terms, the buyer keeps the incumbent, defers the project, or lets it quietly die.

This matters because no-decisions have a different cause and a different fix. They are rarely about price or features; they are usually about perceived risk, urgency or internal momentum. If your win-loss review only studies competitive losses, you will miss the largest category and misdiagnose the rest.

Three practical steps:

• Count no-decisions separately. Report them as their own category so they cannot hide inside “lost.”
• Look for risk signals. Generic demos, unclear implementation paths and weak differentiation all increase the buyer’s fear of making the wrong call.
• Fix buyer confidence, not price. Clearer proof, reference calls and a crisper path to value reduce no-decision rates more reliably than discounting.

Setting your own target: a monthly review rhythm

Benchmarks tell you where the market sits; they do not tell you what your win rate should be. The useful comparison is against your own history and your own segments. A short monthly review keeps the number honest and turns it into a decision tool.

• Report the four numbers – pursuit rate, shortlist rate, win rate and competitive win rate – for the trailing quarter, by segment.
• Compare to your prior quarter, not to a headline. Trend direction matters more than the absolute figure.
• Review the top two loss themes and confirm each has an owner and an action.
• Watch cost per win, so a rising win rate that comes from a spike in bid volume does not hide a worsening unit economics.
• Keep it to 30 minutes. A review that becomes a reporting marathon gets cancelled within a quarter.

The discipline is what produces the result. Programs that run consistently for two years see the lift; one-off studies see nothing.

Pipeline governance: the lever that moves win rate fastest

If you want to move win rate this quarter, the fastest lever is not rewriting. It is bidding less. A significant share of losses is decided before a proposal is written, by pursuing opportunities you were never positioned to win.

• Disqualify earlier. A disciplined go/no-go process improves win rate mechanically, because the bids you decline were the ones dragging the average down.
• Right-size the pipeline. Weak pipeline coverage tempts teams to bid on anything; fixing coverage upstream reduces that pressure.
• Track pursuit rate deliberately. If you are responding to more than roughly half of what you receive, you are likely bidding on low-probability work. In 2026, teams reported responding to about 55% of the RFPs they received (Loopio).
• Separate incumbent and new business. They have different win rates and warrant different investment.

This is the same lever as the go/no-go decision inside the response process; see How to Respond to an RFP.

Build versus buy: in-house and outsourced proposal support

Once you know your numbers, the question becomes how to produce more quality responses without breaking capacity. Three common models:

• In-house, with a governed library and process. Highest control and best long-term economics, but it depends on protecting bid time from daily work.
• Outsourced production, in-house ownership. A human-verified partner drafts and manages the process while you own strategy, pricing and approvals. Fast to start and useful for peak volume.
• Hybrid. In-house for strategic and competitive bids; outsourced for overflow and standard questionnaires.

Choose based on cost per win, not cost per proposal – the model that wins more of the right deals is usually the better one. For the comparison, see In-House vs Outsourced Proposal Support.

A 90-day plan to improve win rate

Measure shortlist rate and requirement coverage monthly throughout. They will move before win rate does, and they tell you whether the fixes are working.

Frequently asked questions

How do you calculate RFP win rate?

Divide proposals won by proposals submitted, then multiply by 100. Exclude undecided bids, and report the number by segment – deal size, industry, incumbent versus new business, and competitor – because a blended figure hides the pattern that matters.

What is a good RFP win rate?

Across industries the average has run around 45% over 2019-2026, with recent averages nearer 39-40% and top-performing teams at 50-60% or higher (Loopio, 2026). New-business bids typically win far less than incumbent renewals, so compare like with like.

Why is my CRM loss reason unreliable?

Clozd’s research found CRM loss reasons disagree with buyer interviews more than 60% of the time, and that a different competitor is identified in roughly seven of ten deals. Buyers are the reliable source; run structured win-loss interviews through a neutral party.

How many win-loss interviews do we need?

Roughly 10-15 per quarter is a practical floor for a lean team, covering a balanced mix of wins, losses and no-decisions. Fewer than eight produces anecdotes rather than patterns.

Does a win-loss program actually improve win rate?

Consistently, when it runs long enough. Clozd’s State of Win-Loss analysis found 84% of programs running two years or more reported a win-rate increase; one-off studies rarely change anything.

What does a proposal cost?

Benchmarks put a single response at roughly 25-41 hours of work and about $2,000-$10,000 in loaded labor, higher for mid-market teams and compliance-heavy bids (APMP and Loopio benchmark data). Track cost per win, not just cost per proposal.

How do you improve win rate quickly?

The fastest lever is selectivity. Disqualify earlier and stop bidding on opportunities you are not positioned to win; tightening bid selection improves win rate mechanically because the bids you decline were dragging the average down. Rewriting proposals is slower to show an effect.

What is a good shortlist rate?

High-performing teams report a median shortlist rate around 63% (AutoRFP.ai, 2026). If your shortlist rate is low, the issue is usually in the written response – compliance, differentiation or structure – rather than the pricing conversation.

Next step

The fastest way to improve your win rate is to define the number properly, tighten what you bid on, and learn from every outcome. Start with the Win-Rate Calculator to set your baseline and segment your results – then book a Win-Rate and Pipeline Review, where we will read your last outcomes, find the two levers that move your number fastest, and run your next response through a human-verified process.

[Use the Win-Rate Calculator] – and get a Win-Rate and Pipeline Review.

Sources

• Loopio, 2026 RFP Response Trends & Benchmarks Report (1,500+ teams, developed with APMP): average win rate ~45% (2019-2026), recent average ~39-40%, pursuit rate ~55%, 40% of revenue tied to formal bids.
• QorusDocs Proposal Management Benchmark Survey: new-business win rate median ~37%; existing-client win rates higher (reported above 70% for many respondents).
• AutoRFP.ai, 2026 Proposal Win Rate Report (94 bid professionals): high-win teams’ median shortlist rate of 63%; win themes and structured process correlate with higher win rates.
• Clozd, State of Win-Loss Analysis: 84% of programs running two years or more report a win-rate increase; CRM loss reasons disagree with buyers more than 60% of the time.
• Corporate Visions, analysis of 100,000+ B2B purchase decisions: sellers and buyers cite different loss reasons 50-70% of the time; 53% of “lost” deals were winnable but for a fixable misstep.
• APMP and Loopio benchmark data (via industry cost analyses): 25-41 hours and roughly $2,000-$10,000 in loaded labor per proposal.

Numbers are cited from their sources and dated. Where a source is a vendor benchmark, it is identified as such. Verify figures against the primary sources before republication.