A single win rate hides where deals die. Stage conversion rates break the funnel apart so a leader can see exactly which transition, which rep, and which segment is dragging revenue down.
A stage conversion rate is the percentage of deals that advance from one pipeline stage to the next within a defined cohort. It is calculated by dividing the number of deals that progressed by the number of deals that entered that stage during the same window. Sales teams use stage conversion rates to pinpoint which transition is leaking revenue, benchmark pipeline health, and prioritize coaching, process, or product fixes.
The six things to understand before you trust a stage conversion number in a forecast review, a QBR, or a board deck.
Stage conversion rate is the share of deals that moved from stage N to stage N+1, measured against the deals that entered stage N during the same cohort window. It is not the share of total open pipeline, and it is not the share of won deals. The denominator is the cohort of deals that got the chance to progress.
A single win rate tells you deals are falling out. Stage conversion rates tell you where. If demo-to-proposal is forty percent and proposal-to-close is seventy percent, the problem lives in the discovery and demo motion, not in the closer. The number is a map, not a scoreboard.
The right denominator is the set of deals that entered the stage during a window, tracked forward to see what happened. The wrong denominator is "deals currently in that stage today," which rewards sitting on open deals and makes conversion look artificially high during slow quarters.
Discovery-to-demo, demo-to-proposal, proposal-to-close all behave differently and should be measured independently. A healthy B2B SaaS funnel often runs discovery-to-demo around fifty to seventy percent, demo-to-proposal around thirty-five to fifty-five, and proposal-to-close around fifty to seventy. Your mix of segment, deal size, and sales motion shifts those numbers.
Counting new business, expansion, and renewal in one bucket produces a conversion rate nobody can act on. Expansion deals convert very differently from net-new logos. Segmenting by deal type, segment, and acquisition source is what makes stage conversion a usable number instead of a vanity metric.
Discovery-to-demo is a qualification problem. Demo-to-proposal is a value and economic-buyer problem. Proposal-to-close is a procurement, pricing, and legal problem. Improving stage conversion means pulling the right lever for the stage that is leaking, not applying one tactic to the whole funnel.
The reason stage conversion rate gets reported wrong more often than any other sales metric is that teams measure it against the wrong denominator. Point-in-time snapshots count whatever is sitting in a stage today, which produces a number that rises when deals get stuck and falls when deals move. Cohort math fixes that by anchoring to a time window and tracking the same group of deals forward. The pattern below is what honest reporting looks like.
Pick a window long enough that most deals in the cohort will finish moving through the stages you are measuring. For a thirty-day sales cycle, a monthly cohort works. For a nine-month enterprise motion, you want a quarter or a half. The window should match the cycle length, not an arbitrary reporting cadence.
The denominator is every deal that entered stage N during the window, regardless of what stage they sit in today. That includes deals that already advanced, deals still sitting there, and deals that were lost or disqualified. If a deal never entered the stage, it is not in the cohort.
For each deal in the cohort, record whether it eventually advanced to the next stage or exited through lost, disqualified, or closed-won from an earlier stage. The conversion rate is advanced divided by the full cohort. Deals still open and sitting in the stage count against conversion until they move.
Stage conversion rate for N to N+1 equals the count of cohort deals that reached stage N+1 or later, divided by the count of cohort deals that ever entered stage N. Keep both counts tied to the same cohort window. Report it as a percentage with the window and segment labeled next to it.
If the average sales cycle is sixty days, do not grade a thirty-day-old cohort. The conversion number will look worse than reality because the deals are still working. The cleanest reporting shows closed cohorts (fully aged past the cycle length) separately from in-flight cohorts still accumulating outcomes.
A single company-wide stage conversion rate hides every actionable insight. Break it out by segment (SMB, mid-market, enterprise), source (inbound, outbound, partner, event), product line, and owner. The segmented view shows which combination is lifting the average and which is dragging it down.
Benchmarks are directional, not prescriptive. The numbers below are typical ranges a B2B SaaS team working mid-market deals might see with a healthy process. Your market, deal size, and sales motion will shift them. The point is not to hit a specific number, it is to notice when a transition sits far outside its typical range and to understand what that gap usually means.
This transition measures qualification. Low numbers mean the top of the funnel is bringing in prospects who do not fit, or that discovery is not surfacing enough value to earn the demo. High numbers (above eighty percent) often mean qualification is too lenient and unqualified deals are polluting later stages.
This is where product-fit and economic-buyer alignment get tested. Low conversion often means demos were technical browses with no decision-maker in the room, or that the value proposition did not survive the second meeting. High conversion sometimes means proposals are being sent without real qualification, which hurts the next stage.
By this stage the fit is proven. Losses here are usually procurement, pricing, security, or legal. A low number means the sales motion is not navigating the buying committee or the procurement process. A high number means the earlier stages are doing their job and the proposal is being sent to the right deal at the right time.
The leap from marketing-qualified lead to sales-accepted opportunity. Inbound inquiries sit in the fifteen-to-thirty range in most B2B teams. Outbound-sourced leads tend to convert higher once they clear SDR qualification, because the targeting is already tighter.
The composite number from a sales-qualified lead through to a sent proposal. This is the number a VP of Sales uses to size pipeline-to-quota coverage. A composite below twenty percent usually signals a qualification problem upstream, not a closer problem.
End-to-end win rate from the first qualified opportunity to closed-won. The product of all the stage conversions in between. Comparing overall win rate to the product of your stage conversions is a sanity check: if the math does not line up, deals are moving around stages in ways the data model is not capturing.
Stage conversion is one of the most commonly reported and most commonly misreported metrics in sales. The errors below show up in nearly every pipeline review where the number is being pulled from a spreadsheet instead of a cohort report. Each one makes conversion look better than it is, which is why forecasts built on these numbers tend to miss on the downside.
Using "deals in stage today" as the denominator rewards stuck deals. The number of deals sitting in a stage grows when conversion is bad, which inflates the ratio against the ones that moved. The honest denominator is every deal that ever entered the stage during the cohort window.
A new-logo deal and an expansion into an existing account behave nothing alike. Expansion deals usually convert at two to three times the rate because the trust and procurement work is already done. Blending them in one funnel produces an average nobody on the team can actually act on.
Inbound, outbound, event, and partner-sourced deals convert at radically different rates at each stage. Reporting one blended conversion rate hides which channel is actually producing revenue. Marketing then takes credit for the wrong channels, and sales invests effort in the wrong ones.
A sixty-day sales cycle produces a terrible conversion number if you grade a thirty-day-old cohort. The deals are still moving. Closed cohorts (fully aged past the cycle length) tell the truth. In-flight numbers are an early indicator at best, a misleading signal at worst.
When reps skip a stage, cohort math breaks unless the reporting handles it. A deal that jumps discovery and lands in proposal did not convert through discovery to demo, it bypassed the measurement entirely. Honest reporting treats skipped stages as implicit progression so the denominator stays whole.
An "all-time" stage conversion rate blends last year's process with this year's. If the sales motion changed, the ICP shifted, or a new product launched, the historical cohort is not telling you about the current one. Every conversion number needs a window stamped on it.
Each stage transition has its own set of root causes, and each root cause has a different lever. A single improvement project aimed at "lift the funnel" rarely works because it mixes qualification problems with pricing problems with legal problems. The playbook below splits the levers by transition so the fix matches the leak.
If this transition is low, the top of the funnel is bringing in misfits. Review the fit criteria, raise the bar on scoring, and give SDRs a disqualification quota alongside the booking quota. If the transition is too high, discovery is rubber-stamping, and demo conversion downstream will suffer.
Demos that end without a confirmed next step and an economic buyer identified rarely convert to proposals. Introduce a required mutual action plan, a stage-exit checklist that includes economic-buyer confirmation, and multithreading practice (two to three contacts engaged per account by the end of the demo cycle).
Losses in late stages are usually procurement, pricing, security, or legal. Pull those workstreams forward so they are in motion before the proposal is sent. A security review that starts after the proposal is a thirty-day drag that often ends in a lost deal. Running it in parallel is the fix.
Deals that sit in a stage past the median almost always convert worse than deals that move at normal pace. Setting a maximum time-in-stage alert (deal sits in demo more than fourteen days, flag it) catches stalled deals before they turn into lost deals. Managers coach the specific deals, not the aggregate.
If reps interpret "qualified" differently, the number is noise. Write stage-entry and stage-exit criteria in plain language, put them in the deal form, and require the fields that prove the criteria are met. Half of the conversion improvement in most teams comes from the data getting more honest, not from reps selling harder.
A manager who runs a generic weekly one-on-one gets generic results. Pointing at the specific stage that is leaking and running tactical coaching on it (role-plays, call reviews, deal reviews on the specific transition) is what moves the number. The data tells you which stage to pick.
Most CRMs ship a stage conversion report. Most of those reports use the wrong denominator by default. The checklist below is what separates a conversion dashboard that leaders actually trust from one that produces an argument in every forecast review. Strkr builds these views in by default, and the AI leak signals surface the specific deals that moved the stage number.
The dashboard should let the user pick the cohort window (month, quarter, half) and lock the denominator to deals that entered the stage during that window. Point-in-time should be clearly labeled as a separate view, not the default. Point-in-time is useful for operational boards, not for diagnostics.
Every stage conversion number is more useful when it is filterable. The dashboard should slice by rep, team, segment, source, product, and deal size. The point is to find the one combination where the leak is concentrated, so the fix is targeted instead of applied across the whole org.
Conversion and velocity move together. A stage with low conversion and long time-in-stage is a stuck stage, and the diagnosis is usually a qualification or process gap. A stage with low conversion but normal time-in-stage is a selling problem. Showing both numbers side by side separates the two.
A single cohort conversion number can be noisy, especially in small teams. The trend over the last four cohorts shows whether the number is drifting, stable, or responding to recent changes. The trend line is what tells a leader whether a fix is working or whether the funnel is quietly degrading.
A conversion percentage is only useful if the user can click into the specific deals that did or did not convert. The drill-down should list every deal in the cohort, color-coded by outcome, so the manager can read the deals that leaked and coach from the specific stories, not from the aggregate.
Strkr watches cohort movement and surfaces leak signals the moment a transition starts underperforming its trend, not at the end of the quarter when the damage is done. The alert names the stage, the segment, and the deals that caused the dip, so the fix is concrete instead of a late post-mortem.
Strkr runs cohort stage conversion by default, segments it by owner, segment, source, and product, and surfaces AI leak signals the moment a transition starts underperforming its trend. Pricing is published. The feature pages show exactly what ships today.
Win rate is the end-to-end percentage of qualified opportunities that end up closed-won. Stage conversion rate is the percentage of deals that advance from one specific stage to the next. Win rate tells you the overall outcome of the funnel. Stage conversion rates tell you which specific transition is driving that outcome up or down.
Define a cohort window (a month or a quarter typically), count every deal that entered the stage during that window, then track each deal forward to see whether it reached the next stage or exited through lost or disqualified. The conversion rate is advanced divided by entered. Use closed cohorts (fully aged past the sales cycle length) to avoid grading deals that are still working.
It depends on the stage and the motion. For B2B SaaS mid-market deals, discovery-to-demo often runs fifty to seventy percent, demo-to-proposal thirty-five to fifty-five, and proposal-to-close fifty to seventy. Segment, deal size, and sales motion all shift the ranges. The honest benchmark is your own trailing cohorts, not a public average.
Teams use point-in-time denominators (deals currently in the stage) instead of cohort denominators (deals that entered the stage during a window). Point-in-time looks better when deals are stuck because the denominator inflates. Cohort math anchors to a window and tracks forward, which produces an honest number that leadership can act on.
Each transition has its own lever. Discovery-to-demo is a qualification problem, usually fixed by tightening the fit criteria and raising the SDR bar. Demo-to-proposal is a value and economic-buyer problem, fixed by multithreading and mutual action plans. Proposal-to-close is a procurement, security, and legal problem, fixed by running those workstreams in parallel earlier.
No. Expansion deals typically convert at two to three times the rate of new-logo deals because the trust and procurement work is already done. Blending them produces an average nobody can act on. Keep them in separate pipelines or at least separate reports, and compare each type against its own trailing cohorts.
A good CRM lets a leader toggle between cohort and point-in-time views, segment by owner and segment and source, drill down from the percentage to the specific deals, and track the trend across the last four cohorts. Without those four capabilities, stage conversion reporting produces arguments instead of decisions.
A leak signal is an alert that fires when a specific stage transition is underperforming its trend, before the quarter ends and the damage is permanent. Strkr AI watches cohort movement in real time and surfaces the exact stage, segment, and deals driving the dip, so coaching and process fixes happen while the quarter is still savable.
No sales call, no migration consultant, no four-month implementation. Enter your card, get 14 days of the full Pro tier, cancel any time before day 14 with zero charge. Spin up a workspace, import your CSV, and have something useful before lunch.
