Every stage.
A reason to move.

What is a sales pipeline stage?

A sales pipeline stage describes the current position of an opportunity in a defined buying process. It should help two teammates reach the same conclusion about the evidence available. “Proposal under review” is useful when it means that the buyer has received a specific proposal and agreed how to evaluate it. It is less useful when it means that someone has attached a draft to a record.

An exit criterion is the evidence required to leave a stage. For example, discovery might finish when the team has confirmed the problem, the people involved in the decision and an agreed next step. The criterion does not guarantee a purchase. It explains why the opportunity is ready for a different kind of work.

This guide presents Meibo’s suggested design method and an illustrative business-to-business pipeline. Adapt it to your actual buying process. It is not a claim that every product supports conditional fields, approvals or automatic stage transitions. Start with a documented operating agreement; then verify which parts your chosen CRM can enforce.

An example pipeline, with evidence at every step

The example below uses five open stages and two outcomes. It suits a considered purchase where a team discusses requirements before agreeing scope and terms. A short transactional sale may need fewer stages. A formal procurement process may need additional approval milestones. The number matters less than whether each stage changes the next decision.

Read the exit conditions as a contract between teammates. Someone taking over the deal should be able to find the supporting note, document or buyer confirmation. If the evidence is missing, the stage should be reviewed even when the original owner is confident. This keeps the board useful during holidays, handoffs and management reviews.

Closed won and closed lost are outcomes, not ordinary work queues. Decide what qualifies as a win in your business: an executed agreement, an accepted order or another documented commitment. Keep payment collection and delivery status separate unless your process explicitly makes them part of the sales outcome.

Decide when an enquiry becomes an opportunity

A new contact is not automatically a deal. A newsletter subscriber, a referral and an active buying conversation represent different levels of intent. Define the point at which the team opens an opportunity: perhaps a relevant need, an identified organisation and agreement to explore a purchase. Keep early enquiries visible without assigning fictional deal values.

Qualification should surface unknowns rather than conceal them. Budget, timing, authority and scope can be unknown at first. Record that explicitly. An empty field cannot distinguish “not asked” from “not applicable”, while a guessed answer makes later reporting look more certain than the underlying conversation.

For example, a consultancy receives an introduction to an operations director. That is relationship context. After a call confirms a particular workflow problem and an agreed discovery session, the team creates an opportunity. The contact and company remain the same records; the potential engagement has its own owner, stage, value and expected decision date.

Write exit criteria a colleague can verify

Use three questions for every open stage: what must be true to enter it, what work happens here and what evidence allows it to finish? Name the owner of the decision and where the evidence is recorded. A requirement such as “good conversation” is too subjective. “Buyer has confirmed the problem and accepted a dated follow-up” is easier to review.

Avoid requiring every fact at the earliest stage. If a buyer has not yet involved procurement, forcing a procurement contact into the first conversation encourages placeholder data. Ask for information when it becomes necessary for the next decision. Keep optional context available without turning record creation into a long administrative exercise.

Test your definitions with real historical examples. Give the same anonymised deal notes to two teammates and ask which stage they would choose. Where the answers differ, improve the definition or identify missing evidence. This is a practical calibration exercise, not a statistical validation study.

A stage can move backwards when circumstances change. A previously agreed scope might reopen after a new stakeholder joins. Preserve the reason and date instead of treating backwards movement as a failure to hide. Check how your CRM records repeated entries before using that history in conversion calculations.

Keep stages separate from activities and next actions

“Email sent” describes an activity. It does not necessarily describe a change in the buyer’s position. A team can send several messages without receiving a decision. If every outbound action becomes a stage, the pipeline can appear to progress while the commercial situation remains unchanged.

Keep a dated next action alongside the stage. At “Proposal under review”, the next action might be “Alex to confirm the review attendees by Thursday”. The stage tells the team where the opportunity stands; the next action tells a person what to do. Both are useful, and neither replaces the other.

The same distinction applies to nurture and delivery. A long-term relationship can stay active even after a specific opportunity closes. A won opportunity can link to an onboarding record without moving through a second sales process. This preserves the meaning of sales outcomes and gives delivery its own responsibilities.

Use another pipeline when the process changes

A separate pipeline is justified when stages or decision rules are materially different. For example, a direct service sale may follow discovery, scope and agreement, while a tender requires qualification, submission and an award decision. Using one sequence for both can create irrelevant fields and misleading stage comparisons.

An industry label or account owner alone may not justify another pipeline. If the buying process is the same, a filtered view may be enough. HubSpot’s documentation similarly recommends separate pipelines for processes with distinct stages and describes using one pipeline across teams when their stages are shared [1]. Verify the access controls and view behaviour in your own system.

Before splitting a process, write down how consolidated reporting will work. Decide which outcomes count toward the same target, how currencies are handled and whether the same commercial opportunity can appear twice. Separate boards should not lead to double-counted revenue or several incompatible definitions of “won”.

Do not confuse a stage with a reliable forecast

A stage probability is a model input. Some CRMs use it to calculate a weighted amount: HubSpot documents stage amount multiplied by the assigned probability [1]. The arithmetic can be correct while the probability is poorly calibrated. A default percentage does not establish your team’s actual likelihood of winning.

Start by separating facts from judgement. Stage, proposed value and expected decision date describe the current record. Probability and forecast confidence express assumptions about the future. Keep the basis visible, and avoid applying one percentage to processes with very different sales cycles or buyer behaviour.

As outcome history grows, compare consistent groups of opportunities and investigate whether assumptions are useful. Include losses and unresolved cases appropriately, preserve the observation dates and avoid treating a small sample as a universal benchmark. The metrics guide explains the difference between a pipeline snapshot and a cohort calculation.

Roll out the definitions without breaking the history

Before renaming, merging or removing stages, map the old definitions to the new ones. Keep a dated copy of the previous configuration and inspect automations, integrations, filters and reports that reference stage identifiers. Changing a label may be simple; changing its meaning can affect every report built on it.

Pilot the proposed stages with a small representative set of opportunities. Ask owners to choose a stage, show the evidence and write the next action. Review ambiguous cases together. Once the team agrees the rules, apply the change deliberately and annotate any reporting period that spans two definitions.

Download the worksheet to document entry conditions, exit evidence and exception handling. Include an owner for future changes. Review the definitions when the buying process changes, rather than adding a new stage for every unusual deal. A useful pipeline remains understandable as the team grows.

Common questions.