Every B2B SaaS company has lead lifecycle stages. Very few have a lead lifecycle model — stages with written entry and exit criteria, agreed recycling paths, and two teams who mean the same thing when they say “SQL”. The difference sounds academic until you try to answer a simple question like “how many qualified leads did we create last quarter?” and get three answers from three systems.
This is the lifecycle model I’ve implemented and refined across nine years in B2B SaaS ops teams. It’s deliberately boring. Lifecycle design is one of those areas where innovation is usually a mistake — the value is in the rigour, not the creativity.
The seven stages, with the criteria that make them real
A stage without written entry criteria is a label, not a stage. For each of these, the criteria must be observable in your systems — a field value, a behaviour, a rep action — or they can’t be automated or reported on.
1. Subscriber. Knows you exist; hasn’t indicated fit or intent. Entry: newsletter signup, single content download. This stage exists mainly so the next one means something.
2. Lead. Identifiable person with minimal engagement. Entry: form fill with enough data to evaluate fit. Exit criteria matter more here: a lead that stays a lead for 90+ days should recycle to nurture or be archived — perpetual leads are where databases silt up.
3. Marketing Qualified Lead. Meets the written, signed fit-and-intent definition — the one covered in depth in the MQL→SQL piece. Entry: crosses the scoring matrix threshold or takes a hand-raise action. This is the stage where drift does the most damage, so it’s the one that most needs a review cadence.
4. Sales Accepted Lead. Often skipped; shouldn’t be. SAL is the stage that makes the handoff measurable — sales has looked at the lead and agreed it meets the definition. Entry: explicit accept action within the SLA window. The MQL→SAL rate is your definition-health metric; without this stage you can’t compute it, and marketing and sales will argue with anecdotes instead.
5. Sales Qualified Lead. Sales has engaged and confirmed genuine opportunity potential — conversation had, need and timeline sketched. Entry: rep-confirmed qualification criteria (pick a framework — the specific one matters less than using one consistently).
6. Opportunity. An open deal with amount and stage. Entry: opportunity record created and associated. From here your pipeline reporting takes over.
7. Customer. Closed-won. The lifecycle doesn’t end — expansion and advocacy are lifecycle stages too — but that’s a different article.
Recycling: the part everyone forgets
Most lifecycle diagrams only flow forward, which means rejected and stalled leads have nowhere to go — so they pile up in stages they no longer belong to, corrupting every conversion rate downstream.
Define the return paths explicitly: rejected MQLs go back to Lead with a captured rejection reason (the reasons are diagnostic gold); stalled SQLs recycle to nurture with a re-entry rule (e.g. new high-intent action within 60 days re-triggers MQL); closed-lost opportunities get a timed re-engagement path, not a memorial in the CRM. Every backward move keeps its history — lifecycle stage should never be edited destructively, or your cohort reporting dies.
Implementing it in HubSpot without breaking reporting
- Use the native Lifecycle stage property as the spine; resist the temptation to build a parallel custom property, which forever forks your reporting.
- HubSpot only moves lifecycle stage forward by default — recycling needs deliberate workflow design (clear-and-reset patterns) and should be one of the few places automation is allowed to write the field.
- Restrict manual edits. Lifecycle stage set by hand, by many people, with no criteria, is how you got here.
- Timestamp every stage entry (HubSpot’s stage date-entered properties, or workflow-stamped custom dates). Stage velocity — time spent in each stage — is where stall points become visible.
- If Salesforce is in the mix, the lifecycle stage sync mapping needs real care; field mapping is its own minefield.
When a lifecycle gets rebuilt, the interesting finding is usually a single stage hiding a stall rather than a funnel that’s slow everywhere. Records pile up in one stage — often an ambiguous “engaged”, or a limbo between MQL and SQL — because the entry and exit criteria were never defined, so nothing ever cleanly leaves. Stage-velocity data makes it obvious: every other stage has a sane median time, and one has a long tail of records that entered months ago and never moved. Define what has to be true to leave that stage and the pile either converts or gets disqualified — either way it stops inflating the pipeline.
The sign-off is the feature
The model above is not novel — deliberately. What makes a lifecycle work isn’t the diagram; it’s the one-page document where every stage’s criteria are written down, dated, and signed by both marketing and sales leadership, with a quarterly review alongside the MQL definition and routing audit. The stages are plumbing. The signature is the product.
—
Lifecycle design is the heart of lens two in the Pipeline Leak Audit. If your stages exist but the criteria live in tribal memory, that’s exactly the kind of thing a free health check can pressure-test in 30 minutes.