Win Rate by Sales Stage in SaaS: Find the Broken Stage Before It Kills Your Pipeline

What Win Rate by Stage Actually Means (vs. What Most Founders Track)

Most founders track one number: closed-won deals divided by total opportunities created. That number is almost useless for fixing anything.

A 22% blended win rate tells you nothing about where you're losing. It's the CRM equivalent of a patient saying "I feel bad" and a doctor prescribing something without running a single test.

Win rate by sales stage measures something different: the percentage of opportunities that move from one stage to the next. Each stage has its own conversion rate, its own failure modes, and its own fix. When you track them separately, a broken pipeline stops being a mystery and becomes a specific address.

The formula for each stage is simple:

Stage Conversion Rate = (Deals that exited Stage N to Stage N+1) / (Deals that entered Stage N)

Stalled deals, those sitting in a stage for more than 30 days with no activity, should be counted as losses for this calculation. If you exclude them, you're flattering yourself.

This is not a theoretical distinction. When Salesforce published its 2024 State of Sales report (surveying over 5,500 sales professionals), teams that tracked pipeline conversion by individual stage were 28% more likely to hit quota than teams tracking only overall win rate. The diagnostic specificity changes the action you take.

---

Benchmark Conversion Rates by Sales Stage in B2B SaaS

These are ranges observed across B2B SaaS companies in the $1M-$10M ARR range. Your numbers will vary by ACV, sales motion (PLG vs. sales-led), and ICP tightness. But the floor matters: below these thresholds, you have a stage problem, not a market problem.

Stage	Healthy Range	Broken (Below)
MQL to SQL	25%-40%	Below 20%
SQL to Demo/Discovery	50%-70%	Below 40%
Demo to Proposal	45%-65%	Below 35%
Proposal to Closed Won	30%-50%	Below 25%

A few things to notice. The SQL-to-Demo rate looks high to founders who've never tracked it explicitly, because they assume every SQL books a call. They don't. An SDR or founder who isn't following up within 24 hours of a SQL trigger routinely drops this number below 40%.

The Proposal-to-Close rate is where most advice focuses. But it's rarely the primary leak. Fixing your close rate when your Demo-to-Proposal rate is 28% is like patching one leak in a boat that has five holes.

---

How to Spot a Broken Stage in 30 Seconds

Pull up your CRM. You need two things: a pipeline view with stage columns and a time filter for the last 90 days. If you have fewer than 20 opportunities in 90 days, use 180 days, sample sizes below 20 produce noise, not signal.

The 30-second method:

1. Count deals that entered each stage in your window.
2. Count how many moved forward (not sideways, not stalled).
3. Divide. Any stage below the "broken" threshold in the table above gets a flag.

In HubSpot, go to Reports > Sales > Deal Funnel report, set the date range to 90 days, and group by pipeline stage. The conversion percentages are pre-calculated. In Salesforce, the standard Pipeline Inspection view does the same job.

Stalled deals are the trap. HubSpot counts them as still "active" unless you mark them lost. Go into your pipeline and look at any deal with zero activity for 21+ days. Those are losses. Mark them as such before you read your funnel report, or your Demo-to-Proposal rate will look 10-15 percentage points better than it actually is.

What stalled deals tell you vs. lost deals: lost deals with a specific reason (price, competitor, no decision) are data. Stalled deals with no close date and no next step are a sales process failure, not a market signal. Don't confuse them.

---

The Stage That Breaks Most SaaS Pipelines at $1M-$5M ARR

The Demo-to-Proposal conversion is where most early-stage SaaS companies quietly hemorrhage pipeline. Not because demos go badly. Usually because the demo goes fine, the prospect seems interested, and then nothing happens.

The reason: no qualification gate between demo and proposal. The founder (or first AE) runs a demo for anyone who books a call, spends 45 minutes on it, and then chases a "let me think about it" for three weeks before the deal dies cold. There was never a real opportunity. There was a curious prospect.

Three signals that this is a process problem, not a product-market fit problem:

Signal 1: Your demo-to-proposal rate is below 35%, but your proposal-to-close rate is above 40%. If your close rate on proposals is healthy but you're writing too few proposals, you're running too many demos on unqualified prospects. The product is fine. The qualification is broken.

Signal 2: Average deal cycle on lost deals is longer than on won deals. You're spending more time on bad fits. That's a pipeline entry problem.

Signal 3: Lost-deal reasons are vague ("not the right time," "budget," "went quiet") rather than specific ("chose Competitor X," "needs feature Y we don't have"). Vague loss reasons mean you didn't qualify, you never knew what the real situation was.

Don't confuse this with PMF problems. PMF problems look like: won customers churning at high rates, ICP that can't be defined, no pattern in who buys. A broken demo-to-proposal conversion at 28% with a healthy close rate above 40% is a pure process problem. You can fix it in 30 days.

---

How to Fix Each Stage (Without Hiring a Sales Ops Team)

MQL to SQL: Tighten the Definition, Not the Volume

If MQL-to-SQL is below 20%, the usual culprit is a loose MQL definition. Every form fill, every webinar registrant, and every free trial signup is getting routed to a "SQL" bucket that isn't actually sales-ready.

Fix: define SQL by three hard criteria: company size matches ICP, the person has budget or influences budget decisions, and they've taken a high-intent action (booked a demo, started a trial, requested a pricing page). Re-score your last 60 days of MQLs against these criteria. You'll likely find that 30-40% of your "SQLs" don't qualify.

Demo to Proposal: The Single-Question Gate

Before you prep a proposal for anyone, get a confirmed answer to this question: "If this solves the problem we talked about, do you have the budget and authority to move forward in the next 60 days?"

Not a soft "yeah we're looking at this." A yes or no. If you can't get that answer on the discovery call, you don't write a proposal. This alone will raise your Demo-to-Proposal rate by cutting junk, and it will raise your Proposal-to-Close rate by improving the quality of what makes it through.

Proposal to Close: Proof of Concept Over Discounting

If your close rate is below 25%, the default response is discounting. This is almost always wrong. Prospects who need 20% off to say yes were not sold on value. They were sold on price, and they'll churn at the same rate.

Instead, offer a scoped proof of concept: 30 days, one specific use case, defined success criteria agreed on paper. Customers who go through a structured POC close at higher rates and with lower churn. Gong's 2024 revenue research found that deals with a formalized POC step had 31% higher close rates than deals where the buyer moved directly from proposal to decision.

What MorBizAI Automates in the Pipeline Feed

A broken Demo-to-Proposal conversion is a sales process problem. But a thin top of funnel makes every fix harder, because you're working from a small sample. MorBizAI keeps content-sourced pipeline consistent: four SEO-targeted posts a month, drafted in under 90 seconds each, with structured CTA paths that hand off warm inbound to your pipeline. You fix the stage conversion rates. MorBizAI makes sure there are enough deals in the funnel to measure.

---

Setting Up a 30-Day Win Rate Audit

You don't need a sales ops hire or a BI tool for this. You need three reports and a tagging convention.

Report 1: Pipeline funnel by stage for the last 90 days. Run it in your CRM today. Screenshot it. This is your baseline.

Report 2: Average days in stage by deal outcome (won vs. lost). Any stage where lost deals sit 2x longer than won deals is a qualification problem, not a closing problem.

Report 3: Loss reason distribution. If more than 40% of your loss reasons are "unknown" or blank, you can't trust any other metric. The first 30 days of the audit should include a requirement that every closed-lost deal gets a tagged reason.

On sample size: if you have fewer than 20 deals per stage in 90 days, your percentage-point swings are statistically meaningless. Don't optimize stage conversion when n=8. Focus on volume first, get to 20+ SQLs per quarter before you start obsessing over whether your Demo-to-Proposal rate is 38% or 43%.

On when to trust the data: 90 days minimum, 120 days preferred. A bad month looks like a trend when the window is 30 days. Founders make the mistake of panic-optimizing off a two-week sample all the time.

Run the audit. Flag the broken stage. Fix that one thing. Then re-run 60 days later and check if the number moved. This is the whole process. It doesn't require a dashboard. It requires discipline.