Most SaaS sales funnels are drawn as a clean pipe: lead in, deal out. In practice it's a series of handoffs between marketing, SDRs, AEs, and the product itself, and each handoff is where deals actually die. Understanding a funnel means understanding conversion rate at each stage boundary, not just top-of-funnel volume or bottom-line close rate — the stage where you're leaking prospects tells you exactly what to fix, and it's rarely "get more leads."
The stages, and what they actually mean
An MQL (marketing qualified lead) is someone who took an action signaling interest — downloaded a whitepaper, attended a webinar, hit a scoring threshold on the website — but has not been vetted for fit or intent. An SQL (sales qualified lead) has been through some qualification, usually a framework like BANT (budget, authority, need, timeline) or MEDDIC, and a rep has judged them worth active pursuit. An opportunity is a qualified deal with a defined next step, a rough close date, and usually a CRM stage attached. Closed-won is revenue; closed-lost is the stage most companies under-instrument, because nobody wants to spend time tagging why deals died.
The MQL-to-SQL conversion rate is usually the noisiest stage in the funnel because MQL definitions drift — marketing under pressure to hit lead targets will quietly loosen the scoring threshold, which inflates MQL volume while SQL conversion craters. When sales starts complaining that "marketing leads are garbage," check the MQL definition before you rebuild the funnel; it's usually a definitional problem, not a targeting problem.
If marketing and sales don't share a written definition of what makes a lead "qualified," the MQL-to-SQL conversion rate is measuring two different things stitched together, and neither team can act on it.
Where conversion actually breaks down
Public benchmarks for B2B SaaS cluster visitor-to-lead conversion in the low single digits, lead-to-opportunity somewhere in the 15-25% range, and opportunity-to-close anywhere from 15% to 30% depending on deal size and sales motion — these vary enormously by ACV and channel, so treat them as sanity checks rather than targets. The stage most teams neglect is the SQL-to-opportunity handoff: this is where an SDR's qualified lead becomes an AE's active deal, and if the handoff process is a Slack message and a prayer, context gets lost and the AE re-qualifies from scratch, which prospects notice and resent.
A funnel stage can have a fine conversion rate and still be broken if deals sit there for weeks with no activity. Track time-in-stage alongside conversion rate — a ballooning average time-in-stage is often the leading indicator of a conversion problem that hasn't shown up in the rate yet.
Product-qualified leads and PLG motions
Product-led growth complicates the classic funnel by adding a PQL (product-qualified lead) — a free-trial or freemium user who hit a usage signal correlated with likelihood to convert (inviting teammates, hitting a usage cap, connecting an integration). PQLs typically convert to paid at meaningfully higher rates than cold MQLs because the prospect has already gotten value from the product before a human ever talks to them. The practical implication for funnel design is that sales should be triaging and prioritizing by product usage signal, not just by firmographic fit — a small company using the product heavily is often a better SQL than a large company that signed up and never logged in again.
Instrumenting the funnel honestly
The most common funnel-reporting failure is survivorship bias in the dashboard: leaders look at a funnel chart built only from deals that made it far enough to have clean CRM data, which quietly excludes the leads that died early from bad data hygiene. A funnel report is only trustworthy if every lead that enters the top is tracked to a terminal state — won, lost, or disqualified — with a reason code. Without that, the "45% close rate" on the dashboard is really "45% of the leads someone bothered to close out cleanly," which is a very different number.
Wrapping up
A SaaS sales funnel earns its name only when every stage boundary has a shared definition between the teams on either side of it, when time-in-stage is tracked alongside conversion rate, and when every lead is tracked to a terminal outcome instead of quietly disappearing from the dashboard. Fix the stage with the worst instrumentation before you fix the stage with the worst conversion rate — you can't optimize what you can't see honestly.
Independent software engineer in Nairobi specialising in Acumatica customisations, Laravel backends, and tax fiscalisation integrations across East and Southern Africa.