Sales pipeline — what it is and how to build one

What a sales funnel is, in plain terms

A sales funnel is a model that splits the customer journey into consecutive steps and counts how many people move from one step to the next. At the top are everyone who reached out; at the bottom, those who paid.

Take 100 leads. You reached 80 of them, 50 booked a consultation, 35 showed up, and 20 paid. That's the funnel: 100 → 80 → 50 → 35 → 20. Every transition between stages has its own conversion rate, and that's what you can improve.

A funnel has two uses:

1. As a report: where and how many customers are lost, which stage underperforms.
2. As a working tool: each customer is a card in its stage, with an owner and a next step. This is what a sales pipeline looks like in a CRM — a kanban board.

Sales pipeline stages

Pipeline stages depend on the business, but they all follow one principle: each stage is a verifiable event, not the customer's mood. "Booked for a date" is a stage. "Interested" is not.

Example for a private clinic:

Example for a service company or agency:

A lost deal isn't junk — it's valuable information in its own right. If 40% of losses have the reason "too expensive", that's a question about your price or how it's explained. If it's "couldn't reach them", that's a question about response speed.

How to calculate sales funnel conversion rates

A stage conversion rate is the share of customers who move on to the next stage. The formula:

Stage conversion rate = customers at the next stage / customers at the current stage × 100%

Overall conversion rate = paying customers / all leads × 100%

Here's a worked example with hypothetical monthly numbers for a clinic:

Now you can work out which stage is more worth improving. Say the average check is UAH 8,000 (Ukrainian hryvnia).

• If you lift "reached" from 80% to 90% (faster callbacks), with every other rate unchanged you get 101 paying patients instead of 90: +11 patients, +UAH 88,000.
• If you lift "showed up" from 75% to 85% (reminders and appointment confirmation), you get 102 paying patients: +12 patients, +UAH 96,000.

Neither improvement requires a bigger ad budget. That's the power of a pipeline: it reveals the levers inside your process. To connect the pipeline to ad spend, see our guide on patient acquisition cost.

How to build a sales pipeline in 7 steps

1. Reconstruct the real customer journey. Take your last 10–20 deals and write down the steps they went through.
2. Define 5–8 stages. Each one is a verifiable event. Fewer than five, and you can't see where you lose people. More than eight, and salespeople stop moving cards.
3. Set the rules for moving between stages. What has to happen for a card to advance, and who does it.
4. Create a list of loss reasons. 6–10 reasons: "too expensive", "couldn't reach them", "chose another clinic", "not our specialty", "changed their mind", "no availability".
5. Assign a next step for every stage. Every card needs a dated task. A card with no task is a forgotten customer.
6. Move the pipeline into a CRM. In a spreadsheet, a pipeline is a monthly report. In a CRM, it's a live board where every move is recorded automatically.
7. Review the pipeline every week. Look at stage conversion rates, stuck cards and loss reasons.

In 95sales CRM, the pipeline is a drag-and-drop kanban board with a list of loss reasons, a first-call questionnaire and quick tasks. Website leads and calls land in the first column automatically, tagged with their ad source.

Which pipeline metrics to track

Beyond stage conversion rates, a pipeline has several other metrics worth checking regularly.

Conversion by source is the link between the pipeline and marketing. A channel with cheap leads can convert worse to payment, so in the end a customer from it costs more. More in our guide to end-to-end analytics for clinics.

How to find and fix a weak pipeline stage

You find a weak stage by comparing conversion rates with your own previous period and across salespeople, not with "industry benchmarks" — there are no universal benchmarks.

Typical problems and fixes:

• Low "lead → reached" conversion. Slow callbacks, missed calls, leads that never reach the front desk. Fix: automatic lead capture, missed-call control, a callback rule. See missed calls.
• Low "reached → booked" conversion. The problem is the conversation: the front desk doesn't uncover the need, doesn't offer a time slot, doesn't handle the price objection. Fix: a script and call recording reviews.
• Low "booked → showed up" conversion. No-shows. Fix: appointment confirmation, reminders, a clear address and time.
• Low "showed up → paid" conversion. The problem is the doctor's consultation or the price. Fix: review loss reasons, installment plans, treatment courses.

Call reviews are especially useful for the second point. In 95sales CRM, AI call analysis scores every conversation against your script stages and shows where the front desk struggles most often.

Sales pipeline in Excel vs in a CRM

You can run a pipeline in a spreadsheet, but a spreadsheet only answers "what happened", while a CRM also answers "what to do now".

A spreadsheet works when you get a few leads a week and one person handles them. Once you have more than one channel or more than two salespeople, a manual pipeline starts drifting away from reality. More on this choice in CRM vs spreadsheets.

Example: a rehab center pipeline

Pipelines differ not only in stages but also in length. Say a 30-bed rehab center works with the relatives of stroke patients. The decision takes one to three weeks, not a day, and it's often made by someone other than the person who called first.

The stages of such a pipeline: lead → first call with a questionnaire (condition, mobility, decision-maker, budget) → tour or online consultation → "waiting for a bed" → contract → admission. A separate "waiting for a bed" stage is needed because there may be no bed available right now, and you can't afford to lose that patient.

Two features of this pipeline:

1. A long cycle. What matters most isn't first-call conversion but making sure every card has a dated next step.
2. Multiple stakeholders. The record needs to capture who makes the decision and who pays — otherwise a salesperson can spend weeks talking to the wrong person.

How this works in an industry CRM is covered on the CRM for rehab centers page.

How often to revisit your pipeline

A pipeline isn't set up once and left forever. Revisit it when your product, channels or team change, and in any case once a quarter.

Signs it's time to change your pipeline:

• cards routinely "skip" a stage — so that stage is redundant;
• cards pile up in one stage without moving — so the rule for advancing is unclear;
• the loss reason "other" is more common than the rest — your list of reasons is outdated;
• salespeople handle some customers outside the system — the pipeline doesn't describe their real work.

Common mistakes when building a pipeline

1. Stages as the customer's mood. "Thinking it over", "interested" — unverifiable, so cards sit there for months.
2. No loss reasons. You can see customers leaving but not why.
3. Cards with no next step. The customer is "in a stage", but nobody knows who will call them and when.
4. One pipeline for different products. If you sell consultations and inpatient stays, or one-off services and subscriptions, the customer journey is different, and so are the stages.
5. A pipeline used only for reporting. If salespeople don't work in it every day, the numbers in the report won't match reality.

The easiest way to roll out a pipeline together with a CRM is to follow a plan — see our guide to CRM implementation. Terms like lead, conversion and LTV are collected in the glossary.