A funnel shows how many potential customers progress towards a purchase. A pipeline shows which individual opportunities are active, what is holding them up and who must act next. You need both views, but they answer different questions.
Imagine the same business viewed at two different moments. The funnel asks what happened to a group of enquiries over time. The pipeline asks what is happening with each open deal right now. Neither view substitutes for the other.
Count prospects at defined steps, compare conversion rates and examine where opportunities disappear or take longer to progress.
Review individual deals, stage evidence, expected value, decision timing, responsible owner and the next buyer-dependent action.
Terminology varies. Some firms call every step from first enquiry to purchase a “sales pipeline”; others begin the deal pipeline only after qualification. The labels matter less than defining precisely which population and event each report counts.
Suppose a B2B services firm tracks 120 enquiries received in one quarter and follows them long enough to know the eventual outcome of each. Here is the same cohort at successive steps, not a snapshot of whatever happens to be open today.
Enquiry-to-order conversion
Opportunity-to-order conversion
Orders from this cohort
The numbers alone do not tell us whether the process is healthy. Perhaps half the incoming enquiries were never suitable. Perhaps nine orders are enough because contracts are large and profitable. Perhaps prospects who did not request a proposal are still buying later.
Denominator matters: nine orders divided by 120 enquiries measures a different thing from nine orders divided by 30 qualified opportunities. So does a win rate calculated only from closed deals. State the denominator and the observation period alongside every conversion figure.
Timing matters: a quarter's new enquiries and the same quarter's closed orders often belong to different cohorts. Dividing one by the other creates a rate that may move simply because deal cycles are long. Use a defined entry cohort with adequate follow-up, or an explicitly labelled period-based throughput measure.
The 30 opportunities in the example are not necessarily the 30 deals open in your CRM today.
A live pipeline combines opportunities created on different dates. One may have reached proposal this morning. Another may have sat at “proposal sent” for four months. A third may have been reopened after a customer changed its purchasing plan.
For each open opportunity, you need a buyer, a reason the buyer may act, an agreed next step, an owner, an expected value and a view of timing. Merely assigning every opportunity a stage does not make these conditions visible.
One pipeline report might show £400,000 in open opportunities. That is neither £400,000 of revenue nor evidence that the business is close to winning £400,000. You must ask which opportunities have a credible path to decision.
These are illustrative stages for a consultative B2B sale. Use fewer or different stages if the buyer's process warrants it. A stage should correspond to a meaningful change in what is known about the customer's purchase decision.
| Stage | Evidence required to enter | Question for the owner |
|---|---|---|
| 1. Qualified opportunity | A credible need, buyer or buying group, fit and a plausible route to a decision have been established. | What decision might the customer actually make? |
| 2. Discovery completed | Requirements, current situation, decision participants and significant constraints have been checked with the buyer. | Which uncertainties remain before a solution can be proposed? |
| 3. Solution evaluation | The buyer is actively reviewing an approach or scope; the criteria and next evaluation step are understood. | What must the customer accept or verify? |
| 4. Proposal under decision | A specific proposal has been provided to the relevant buyer and there is an understood decision process or next step. | Who can approve, reject or defer it — and when? |
| 5. Contracting | The buyer has indicated intent to proceed, subject to identified contractual, procurement or approval conditions. | Which condition remains unresolved? |
| Closed won / closed lost | An agreed sale is confirmed, or the opportunity has definitively ended. Record outcome and reason. | What must transfer to Delivery, and what did we learn? |
Important: scheduling a meeting or sending an email is an activity, not automatically evidence that a buyer has progressed. Likewise, “proposal sent” is not always the same as “proposal being evaluated”.
Each stage needs rules for what happens when a deal stalls, is disqualified, returns to an earlier stage, or is postponed without being lost. Decide whether renewals and expansions are separate opportunities, and how one account can hold several simultaneous buying decisions.
Number and source of enquiries, qualified leads and opportunities. Always examine the customer mix, not only the total.
Transitions for a defined cohort, lost and disqualified reasons, and conversion split by source, segment or offer.
Time between entry and exit, stalled deal age, decision-date changes and the share of work awaiting buyers versus internal teams.
Order value and expected contribution, delivery constraints, contract terms and the cost of acquiring and serving a customer.
A falling overall win rate might result from a larger share of cold enquiries rather than worse sales execution. A rising win rate might result from qualifying out difficult buyers earlier. A faster sales cycle might reflect smaller, simpler deals. Review the process and the mix together.
CRM systems can assign a probability to a deal stage and calculate a weighted value. HubSpot, for example, lets teams configure stage probabilities. That can be useful for modelling, but the multiplication does not validate the underlying forecast.
A £100,000 opportunity at a nominal 60% stage probability becomes £60,000 of weighted pipeline. It does not establish a 60% chance of that particular customer buying in the forecast period.
Compare past outcomes for similar deals, the buyer's actual decision steps, slippage and contract timing before using stage probabilities to make staffing or cash decisions.
Before buying another dashboard or changing your CRM, try to trace a small sample from first enquiry to the actual decision.
If you cannot reconstruct this from current records, a reported funnel conversion or forecast percentage may be much more precise than the underlying information.
Marketing can be accountable for generating suitable enquiries. Sales can be accountable for qualification, buyer engagement and opportunities. Delivery can be accountable for fulfilling the order. But someone must define the handoffs and reconcile what each function calls success.
An opportunity may be closed won even though implementation is delayed, the customer asks for additional work, or the invoice will be recognised later. Pipeline value, booked orders, recognised revenue and profit therefore need different definitions and responsible owners.
See what can be lost at the Sales–Delivery handover and how the management view connects commercial and operational records.
No. A lead may be a contact or an enquiry. An opportunity represents a potential buying decision with some identifiable scope. Different CRMs use these objects differently; document your rules before reporting conversion.
Usually not. Many businesses first assess fit and intent, then create an opportunity when there is a credible commercial decision to pursue. Otherwise deal counts and forecasts can be inflated by contacts who are not buying.
Yes, if opportunities, stage evidence, owners and outcomes can be reliably maintained. CRM becomes more useful when records and handovers cannot be managed consistently by an individual or a small team.
No. A funnel is a business measurement of progression and conversion. A customer journey includes what the buyer is trying to accomplish, their alternatives, evaluation and experience — not all of which are captured by your internal sales stages.
That is common in consultative services. Buyers may repeat evaluations, defer decisions, or purchase several services separately. Stage definitions still help if re-entry, postponement, multiple opportunities and outcomes are recorded explicitly.
Separate suitable demand from progression, timing and offer fit.
Explore sales and enquiries →Consider whether customer, deal and delivery records answer different questions.
Examine CRM record design →Test stage evidence, assumptions and missing decision events.
Examine forecast uncertainty →Further reading: HubSpot — Sales pipeline stages and funnel comparison; HubSpot Knowledge Base — configuring deal stages and probabilities. The examples and framework on this page are Unsnag's own analysis; no HubSpot affiliation is implied.