Why stages need written definitions
Ask three people in the same company what a marketing-qualified leadA lead that marketing judges ready for sales follow-up against agreed criteria. is and the answers will differ, and usually none of them will be written anywhere. One means a lead that filled in a form; one means a lead with a score above a line somebody set a year ago; one means a lead the sales team did not complain about. Every report of leads by stage in that company is built on all three definitions at once, and the number at the top of the monthly deck is an average of disagreements.
Stage definitions fix this by turning stages from labels into tests. A definition says what must be true of a record for it to be in the stage. An exit criterion says what event moves it to the next one. An owner says who is accountable while it sits there. When these three things are written, a stage change becomes a fact that a workflow can act on and a report can count, rather than an opinion a salesperson expressed by dragging a card. The tutorial on automating lifecycle stage changes builds those workflows, with a date recorded as each stage is entered and backward moves allowed only by rule.
The definitions also carry the relationship between marketing and sales. Marketing's commitment is to deliver leads that meet the definition of the stage it hands over; sales' commitment is to work those leads within an agreed time and to record what happened. Neither commitment can be kept or checked while the stage it refers to is undefined, which is why the argument about lead quality in most companies is really an argument about a missing document.
The customer journey stages in the journey stages guide describe what a person is doing at each point; the pipeline stages here describe what the business has established about them. The two line up but are not the same thing, and the pipeline is deliberately the narrower, more testable set.
The stages and exit criteria
The stage names below are the common ones. The names matter less than the fact that each has a definition, an exit criterion and an owner, and that the three are agreed by both teams and stored where both can see them.
| Stage | Definition (what must be true) | Exit criterion (what moves it on) | Owner |
|---|---|---|---|
| Subscriber | Has given consent to hear from the business but has not asked for anything | Submits a form or takes an action that signals interest | Marketing |
| Lead | Has identified themselves and expressed interest in something the business sells | Meets the fit and intent threshold agreed with sales | Marketing |
| Marketing-qualified lead (MQL) | Meets the written fit and intent threshold; not yet reviewed by sales | Sales reviews and accepts or rejects within the service level | Marketing, pending sales |
| Sales-accepted lead (SAL) | Sales has reviewed the record and agrees it is worth a conversation | First conversation takes place and need, authority, budget and timing are explored | Sales |
| Sales-qualified lead (SQL) | A conversation has confirmed a real need, a route to a decision and a plausible timeframe | A deal is created with an amount and expected close date | Sales |
| Opportunity | An open deal with a stage, amount and close date | Won, lost, or stalled past the agreed limit | Sales |
| Customer | At least one won deal | Renewal, expansion or churn, handled by lifecycle marketing | Account management |
| Disqualified | Reviewed and found not to fit, with a recorded reason | Recycled to nurture if the reason is timing; closed if the reason is fit | Whoever disqualified |
A few points on the table. The sales-accepted stageA lead sales has reviewed and agreed is worth working, recorded as a stage between marketing-qualified and sales-qualified with a reason when rejected. is the one most companies skip and most companies need, because it separates "marketing sent it" from "sales agreed it was worth sending" and makes the rejection reason a recorded field rather than a hallway complaint. The customer stage is where the pipeline ends and lifecycle marketing begins. And disqualified is a stage, not a deletion: a lead that was wrong on timing is the best lead a nurture sequence will ever get.
Definitions should be testable from fields on the record. "Has budget" is an opinion; "budget field is not empty and is above the minimum deal size" is a test. Where a criterion depends on a conversation, the record of that conversation (an activity with a type and an outcome) is the field.
MQL, SQL and the hand-off
The marketing-qualified to sales-qualified transition is the join between two teams, and it fails in two symmetrical ways: marketing passes too much, and sales works too little. Both are fixed by the same document, usually called a service level agreement, which states what each side owes the other.
Marketing owes a definition and a volume. The definition is the MQL threshold: which combination of fit (does this person and company look like a customer) and intent (have they behaved like one) qualifies. The lead scoring guide covers how fit and intent become a score; the threshold the score has to reach is set here, by the two teams together, and revisited when the pipeline shows it was wrong. The lead scoring model template keeps the agreed threshold, the points behind it and the last calibration on one sheet. The volume is a monthly commitment that sales can plan capacity against.
Sales owes a response and a disposition. The response is a first attempt to contact within an agreed window, logged as an activity. The disposition is an accept or a reject, with a reason from a picklistA CRM field type that accepts one value from a defined list of permitted values, so that records can be reported on instead of typed in many spellings., within a slightly longer window. A lead that is neither accepted nor rejected after the window is a breach on the sales side, and the report that counts those breaches is the one that makes the agreement real.
The hand-off itself should be a workflow. When a record crosses the threshold, the system changes the stage, assigns an owner by the agreed ruleRules that assign each new lead to an owner, queue or next step based on territory, segment, score or source., creates a task with a due time, and notifies the owner. Nothing is sent by hand and nothing waits for a weekly meeting. The lead routing QA checklist tests that workflow end to end with test leads whenever a stage, field or territory changes, and the agreed rule itself is written down one route at a time in the lead routing rules template. For business-to-business teams the hand-off is often where lead-generation forms on professional networks meet the pipeline, and the LinkedIn business-to-business hub describes how sales and marketing share that platform; the pipeline rules are the same whatever the source.
Rejections deserve their own report. If a large share of marketing-qualified leads are rejected for the same reason, the threshold is wrong or a source is poor, and the fix is in marketing. If leads are accepted but never converted, the problem is downstream. Without the reason field the two look identical.
The MQL versus SQL comparison sets the two statuses side by side, including who decides each and how to report the step between them without double counting.
Speed to lead and service levels
How quickly a new lead is contacted matters more than most things sales and marketing argue about, and it is one of the few variables entirely within the business's control. A person who has just submitted a form is at the peak of their interest; every hour that passes without contact lets that interest fade and lets a competitor get there first. In practice the window in which a first contact is most likely to reach the person is short, and companies that measure it are usually surprised by how long their own median is.
Speed to leadAverage time between a new lead arriving and the first contact from sales. is measured as the time between the stage change that created the marketing-qualified lead and the first logged outbound activity by the owner. It needs three things to be measurable: a timestamp on the stage change (which the CRM normally records), a logged activity (which salespeople have to be required to create, or which calling and email tools have to create automatically), and a report that joins the two by lead and summarises by owner, source and time of day.
The service level sets the target. It is usually expressed as a maximum time to first attempt during working hours, with a separate rule for leads that arrive outside them, and a maximum number of attempts before a lead can be dispositioned as unreachable. The same document sets the acceptance window from the previous section. Targets should be set from the current distribution, tightened as the process improves, and written as time windows rather than aspirations.
Automation closes most of the gap. The hand-off workflow assigns and notifies within seconds; an escalation rule reassigns a lead that is untouched past the window; and an automated first-touch email, sent in the owner's name and inviting a reply, buys time without pretending to be the conversation. What automation should not do is log an activity that did not happen, because the speed-to-lead report is only useful if the first attempt it measures was real.
To find out how much speed matters for your own leads, the lead response time test randomises routing at submission and reads meetings held per lead.
Stage conversion reporting
Stage conversion is the share of records entering one stage that go on to enter the next. Measured for every pair of adjacent stages, it shows where the pipeline leaks, and measured by source, by owner and by period, it shows why. It is the pipeline's conversion rate (CVRShare of visits or clicks that convert.), and the conversion rate calculator does the arithmetic for a single pair of counts.
The report has to be built on cohorts. Counting how many leads were created this month and how many deals were won this month gives a ratio of two unrelated populations, because the deals won this month came from leads created in earlier months. The correct report takes the records that entered a stage in a period and follows that cohort forward, so the conversion from lead to opportunity for the leads created in a given month is measured once those leads have had time to convert. Older cohorts are complete; recent ones are still in progress and should be read as such.
Three cuts turn the report into decisions. By original source, it shows which channels produce leads that progress and which produce leads that stall, and it is the basis on which cost per acquisition (CACCost to acquire one new customer.) is calculated at the stage the business actually cares about rather than at the form. By owner, it shows whether a stage's conversion varies with who works it, which is a coaching question rather than a marketing one. By period, it shows whether a change to a threshold, a form or a sequence moved anything. The conversion analytics section covers the same analysis for on-site funnels, and the methods transfer. Events need the same cohort reading, with attendance and follow-up added before the first stage, as the guide to measuring webinars and events describes.
The drop-off view is the simplest way to read it: a bar per stage, shrinking left to right, with the loss between each pair labelled. The explorer below lets you enter counts for each stage and see where the largest loss sits; the largest loss is not always the most fixable one, but it is always the first to understand.
Pipeline drop-off explorer
A software company reads its quarter
A business software company creates 1,800 leads in a quarter. 540 cross the marketing-qualified threshold, sales accepts 410 of them, 150 become sales-qualified after a first conversation, 95 become opportunities and 31 are won. The largest absolute loss is between lead and marketing-qualified, where 1,260 records never reach the threshold; the largest proportional loss is between accepted and sales-qualified, where nearly two-thirds of accepted leads do not survive the first conversation. Cutting by source shows that leads from a comparison-site listing convert from accepted to qualified at 19 percent while leads from webinar registrations convert at 52 percent, which moves the conversation from "sales is not working the leads" to "one source is producing people who are not ready". The comparison-site leads are routed to a nurture sequence instead of straight to sales, and the next quarter's accepted-to-qualified rate rises without any change on the sales side.
Recycling and disqualification
A lead that is not ready is not the same as a lead that is wrong, and a pipeline that treats them alike loses the first kind. Disqualification needs a reason, chosen from a short picklist agreed by both teams: wrong fit (company size, sector, geography), no need, no budget now, timing, competitor chosen, unreachable, and duplicate or junk. The reason decides what happens next.
Fit and junk reasons close the record: the lifecycle stageA CRM field recording where a contact stands in the relationship, such as subscriber, lead, opportunity or customer. moves to disqualified, workflows stop, and the contact is retained only as long as consent and retention rules allow. Timing, budget and need reasons recycle the record: the lifecycle stage moves back to lead, and the record enters a nurture sequence designed for its reason, with a re-entry rule that lets it cross the marketing-qualified threshold again when behaviour changes. Lost opportunities are recycled the same way, with the lost reason doing the work of the disqualification reason.
Re-entry has to be designed, not assumed. A recycled lead that visits the pricing page should be able to become marketing-qualified again, but a lead that was disqualified for fit should not, however many pages it reads, because the fit has not changed. The scoring model handles this with negative or capped scores for closed reasons; the stage definitions handle it by stating which prior reasons permit re-entry. The KPIA metric tied directly to an objective. for the recycling programme is the share of recycled leads that re-qualify and go on to convert, measured by the reason they were recycled for, and it is usually the cheapest pipeline the business has.
Reporting should show recycled and disqualified records separately from new ones. A pipeline that counts a re-qualified lead as new overstates marketing's new-lead performance and hides the recycling programme's value at the same time.
Stage dates recorded in the CRM are what make cohort reading possible in AARRR for business-to-business lead generation, where each pirate-metrics stage is redefined for a lead that is not yet a customer.
Close rates reported from these stages are the key input in the allowable cost per lead guide for the break-even calculator, which turns them into a maximum price per lead.
If records cross the hand-off line and then sit with no logged activity, the unworked leads diagnosis traces whether routing, a broken sync, thin records or a lead peak left them untouched.
Common mistakes
- Stages without written definitions. Every report built on them averages several private meanings.
- No sales-accepted stage. Marketing cannot tell "rejected" from "ignored", and the lead-quality argument never ends.
- Thresholds set inside the scoring model. The threshold is an agreement between teams; the model is a calculation. Keep them separate so either can change.
- Measuring conversion on unrelated populations. This month's wins over this month's leads is a ratio of two different cohorts.
- Speed to lead measured on claimed activity. If the first attempt is not logged automatically or verified, the report measures diligence in logging, not in calling.
- Deleting disqualified leads. The timing-reason leads are the best nurture audience in the database.
- One pipeline for every product. Stages that mean different things for a self-serve subscription and an enterprise contract cannot share definitions; use separate pipelines with their own documents.
- Never revisiting the threshold. The marketing-qualified definition should change when the conversion report says it is wrong, on a scheduled review rather than after an argument.
The CRM for marketers hub places this guide beside the fundamentals and integration guides, and the lead scoring guide takes the thresholds agreed here and builds the model that reaches them.
