Sales pipeline: what it is, its stages and what to measure

Marcel GarciaUpdated Oct 4, 2026

A sales pipeline is the list of open opportunities, ordered by the stage each one is in, from the first real conversation until they sign or are lost. It tells you how many clients you can expect to close and when, and where you lose the rest.

When you sell something expensive, every opportunity counts. You close few clients a year, the sale takes weeks or months, and each one you lose is a lot of money. Without a pipeline, the only signal is the feeling that "this month is slow", and it arrives late: what you don't sign today is what didn't come in one sales cycle ago. The pipeline turns that feeling into a number you see earlier, as far ahead as your own cycle.

This guide is signed by Marcel Garcia, founder of HAKI, and it is written for businesses that close between ten and fifty clients a year, where the founder usually does the selling. If you have a sales team with quotas, the CRM vendors' manuals will serve you better. The pipeline it shows is ours, the one we use to sell our own projects. The examples with round numbers are just that, examples; the client figures are only the ones a case page publishes.

What a sales pipeline is, and how it differs from a funnel and a forecast

An opportunity is a conversation with someone who can buy from you: they have the problem, they can pay for the solution and they make the decision. Not every email that comes in is one, and not every form: those are leads, and they wait in the inbox until the first call. If you put everything in the pipeline, the total goes up and stops meaning anything.

A pipeline and a funnel are not the same thing. The funnel counts visits, leads and requests: that is marketing. The pipeline counts opportunities and what they are worth: that is sales. The funnel feeds the pipeline, which is why they belong on the same page, but a full funnel with an empty pipeline usually means the wrong people are coming in.

Nor is it a forecast. The forecast is what you expect to close in a period, and it is calculated from the pipeline, for example as each opportunity's value times the close probability of its stage. Without an honest pipeline, a forecast is a wish with decimals.

Sales pipeline stages: the ones HAKI uses

A sales pipeline usually has an inbox, where leads wait, three active stages (discovery call, proposal and closing) and three outcomes: won, unqualified and lost. Ours adds a long-term stage, and it is the pipeline HAKI sells its own projects with, in Notion, not a CRM template. Each stage is defined by something the buyer has done, not by something you did: "email sent" is not a stage.

	What has to have happened to enter	What you measure on the way out
1. Inbox	There is a name, a contact and a sentence on what they sell. You record the source and, if they came through the website, the page they came in on. It is still a lead: it doesn't count in the pipeline until the first call.	The share that reaches a first call, by source.
2. Discovery call	A first call has happened with whoever decides. You know their problem and what an average client spends with them.	The share that gets a proposal, and why the rest are "unqualified".
3. Proposal	Whoever decides has asked for a proposal, on a problem and an amount discussed on the first call.	The share that moves to closing, and how many days each opportunity spends here.
4. Closing	The client accepts the approach and you are negotiating terms or the contract.	The win rate, and the reason when it is lost.
Long-term	There is a fit, but the timing is wrong, and there is a date to talk again.	How many come back to discovery or proposal.
Won, unqualified or lost	Closing it takes the amount if it is won, or the stage it died in and a reason if it isn't.	Where opportunities die, and why.

How to build a sales pipeline (in Notion, a spreadsheet or a CRM)

With a few dozen opportunities a year you don't need a CRM: a table in Notion or a spreadsheet is enough, as long as it records six things and always gets filled in. Who it is and what they sell. Which source they came from and, if they came through the website, which page they came in on. Which stage it is in. When it came in and when the last contact was. What it is worth. And, if it closes without a win, the stage it died in and why. A short table that always gets filled in gives better data than a CRM with forty fields nobody touches.

Define each stage by something the buyer has done, as in the table above, and review the pipeline once a week: what has moved, what has been sitting still too long and what should be marked as lost. If you move to a CRM one day, you take a clean table with you, not a year of loose notes.

How much pipeline do you need? The rule of carrying three times your target in pipeline is meant for sales teams with quotas. With few sales a year, count opportunities, not euros: if you close one in four and want two new clients a month, you need about eight new opportunities a month, not counting the unqualified ones, one sales cycle ahead.

What to measure: the four numbers of a sales pipeline

Four figures tell you how much your pipeline produces and what holds it back: how many opportunities you have, what share you close, what each one is worth and how long they take to close. Each one points at a different constraint, and sales velocity puts them together in one.

Open opportunities

How many there are today from the first call to closing, and which source each one came from. Leads waiting in the inbox don't count, and neither do long-term ones, or the ones that will never sign even though nobody has marked them as lost.

Win rate and conversion between stages

Won deals divided by closed opportunities, won or lost, leaving out the unqualified, which are not a sales failure. With few sales a year, read it over the last twelve months. Split by stage (inbox to first call, first call to proposal, proposal to close) it tells you where to look: the stage that converts worst is usually the one limiting growth, not the noisiest one. It is one of the five numbers of growth marketing.

Average deal value

What a won client is worth on average. Next to CAC it tells you how much you can pay for each client and, with the win rate, for each opportunity. CAC, in turn, needs the pipeline's dates to match spend with the clients it brought in.

Cycle length and days in each stage

How long an opportunity takes to close from the first call, and how long it spends in each stage. A stage that piles up days hides a decision nobody is pushing, or a buyer who has already left and nobody has marked as lost.

Sales velocity: the four in one figure

Open opportunities × win rate × average deal value ÷ cycle length in days. It tells you how much your pipeline produces per day and puts every lever in the same unit, euros a day, so you can see which one is worth moving. Finding the one holding you back is the first step of the method.

An example with round numbers

A business that sells projects of around €6,000, with a two-month sales cycle. The numbers are made up so they come out round; the method is the one we use.

Today: €500 a day

20 open opportunities × 25% win rate × €6,000 ÷ 60 days. That is what the pipeline produces as it stands: five clients over the next two months, about €15,000 a month.

A 45-day cycle: about €667 a day

Same opportunities, same rate, same price. You only remove fifteen days of waiting: a proposal that goes out before the buyer cools off, a follow-up nobody forgets. A third more.

A 30% win rate: €600 a day

Five points more, for example by no longer sending proposals to people whose ticket didn't fit on the first call, or who weren't the ones deciding. 20% more.

With more opportunities, you would need 27

To match the 45-day cycle without changing anything else, you need seven more open opportunities, and whatever it costs to win them. In a long sale, shortening the cycle or fixing one stage is usually worth more than pouring more leads in at the top.

Lost, but at which stage?

It is the piece of data most often missing from the pipelines we are shown, and the one that decides where to work. In ours, every opportunity that closes without a win has two fields for it: the stage it died in and a reason.

The stage it died in

"Lost" is not enough. Losing on the first call says something about who comes in; losing at the proposal says something about the offer or the price. They are different problems, with different fixes.

A reason, from a closed list

Poor fit, lost to a competitor, done in-house, no response, no real pain, price, internal restructuring, shifted priorities, budget eliminated or unreachable. A closed list can be counted; a free-text comment can't.

Unqualified is not lost

Someone who was never going to buy doesn't count against sales. Keeping them apart tells you whether the hole is in acquisition or in closing.

A maximum time without contact at each stage

When an opportunity goes past it, it is flagged for follow-up. What comes next can be automated: for Nuria Marco, a pre-qualification form collects each request and she decides whether she can help; if she can, a flow explains the method and the offer and follows up when there is no reply, without a single call.

Before the pipeline: lead generation, measured by source

Lead generation is everything that fills the first stage: SEO, referrals, outbound, ads, directories. For the pipeline, one distinction matters that almost nobody makes: the source is not the channel. The source is the strategy that brought that person in; the channel is how they got in touch (the form, an email, a call). If you only record the channel, everything seems to come from the form.

We ask every source three questions. Is it warm or cold: did they know you or not? Is it one-to-one or one-to-many? Can you scale it: if you double the input, do the leads roughly double, or does the volume depend on other people, as it does with referrals? With that, and the win rate of each source, you know where to put the next euro.

Since 30 September 2026, every lead that comes in through HAKI's form also records the page they entered the site on. That way a page is judged by the opportunities it produces and by the ones that end up signing, not by its visits.

The mistakes that inflate a pipeline

They make the total look healthy while the business doesn't close.

Counting leads as opportunities

A form is not an opportunity until someone who decides has spoken with you. If you count everything, the win rate collapses and you can no longer tell whether the problem is acquisition or sales.

Never closing anything as lost

The zombie pipeline: opportunities nobody has discarded that are never going to sign. They inflate the total and the forecast, and hide the stage where they died.

Measuring by calendar month

In a two-month cycle, March's signatures come from January's opportunities. Read conversion by grouping each opportunity by the month it came in, or you will compare months that have nothing to do with each other.

Not recording the source

It is recorded on the way in. Reconstructing it months later, when the client signs, is guessing, and without it you don't know which channel to feed.

What it looks like when it works

Both cases are health practices with a short cycle: someone books an appointment, they don't sign a year-long contract. They don't prove a long sale; they prove the habit that makes one measurable: counting appointments, not clicks, and knowing which source they come from. We don't publish any client's pipeline value, their win rate or any figure beyond the ones on their case page: it is their data, not ours.

Laura Jové: measured in appointments

An online health practice with a single call to action: a free 20-minute assessment that books itself. What leads to the appointment is measured: more than 10% of visits open the calendar, and in the ads the conversion is the booked appointment, not the click. And appointments are counted by source: from Google Ads, 246 in 2025 and 339 from January to September 2026; from organic search, 51 in 2025 and 151 so far in 2026.

Clínica Albareda: measured by source

A traumatology and regenerative medicine clinic in Barcelona. Every page ends in the same place, booking an appointment, and the form is measured as a conversion. That is how we know where requests come from and how many there are: 1,048 appointment requests from Google in 2025, 82% of all of them and with no advertising; in 2022 there were 223.

Keep going with the other guides

CAC

What it costs to win a client, how to calculate it per channel and what a good LTV:CAC ratio is.

Go-to-market

How to take an offer to market: who, which problem, at what price, through which channel and with what proof.

Lead scoring

How to tell who can buy from you from who can't, before the call.

Selected projects

Embat

Fintech, Tech & SaaS

Clínica Albareda

Health

Laura Jové

Health

View all projects

A growth partner is paid partly on what the pipeline closes, not on leads

A base plus a variable part tied to new clients, sized so the total stays between 10% and 25% of the new revenue. In the variable part, leads don't count: what signs does. What it is, what it costs and when it fits.

What a growth partner is

An opinion, signed

The most useful number in a pipeline is not its total value: it is the stage where each opportunity dies, and why. A founder who closes twenty clients a year doesn't need a weighted forecast: closing five out of twenty opportunities a quarter, a single deal won instead of lost moves the win rate by five points. They need the lost-at-stage column. If deals die on the first call, the wrong people are coming in and the problem is acquisition; if they die at the proposal, the problem is the offer or the price, and no extra lead will fix it.

For the same reason, a page that brings ten contacts who die on the first call is worth less than one that brings two who sign. That is why every lead that reaches us through the form carries the page it entered the site on, and why I will judge this guide by the opportunities it brings, not by its visits.

Marcel Garcia, founder of HAKI. Growth partner to businesses that sell on value, from Barcelona.

Contact

Tell us what you sell and where your deals get stuck.

hello@haki.marketing

Name

Surname

Phone

Company

Role

What do you sell?

What does an average client spend with you?

Who decides the marketing budget?

Message

Sales pipeline questions

It is the list of your business's open opportunities, each in the stage it has reached, from the first conversation with someone who can buy until they sign or are lost. It lets you estimate how many clients you will close and when, and see where you lose the rest.

The funnel counts visits, leads and requests; the pipeline counts real opportunities and what they are worth. The funnel is marketing and feeds the pipeline, which is sales. A full funnel with an empty pipeline usually means the people coming in are not going to buy.

The pipeline is what you have open today; the forecast is what you expect to close in a period, calculated from the pipeline, for example by multiplying each opportunity's value by the close probability of its stage. If the pipeline is inflated, so is the forecast.

It usually has an inbox, where leads wait, three active stages (discovery call, proposal and closing) and three outcomes: won, unqualified and lost. You can add a long-term stage for opportunities with a fit but not the right timing. The fewer the better, as long as each one means something: what matters is that it is defined by something the buyer did, not by a task of yours.

With a table that records six things and always gets filled in: who it is and what they sell; which source they came from and, if through the website, which page they came in on; which stage it is in; when it came in and when the last contact was; what it is worth; and, if it closes without a win, the stage it died in and why. Define each stage by something the buyer has done and review it once a week.

With few sales a year, count opportunities, not euros. Divide the clients you want by your win rate: if you close one in four and want two new clients a month, you need about eight new opportunities a month, not counting the unqualified ones, one sales cycle ahead. The three-times-target rule is meant for sales teams with quotas.

Whichever one your team always fills in. While one or two people do the selling, a spreadsheet or Notion is enough; ours is in Notion. A CRM such as HubSpot or Pipedrive pays off when several people are selling or the volume can no longer be logged by hand.

It is how much money your pipeline produces per day. Multiply open opportunities, from the first call to closing, by the win rate and by the average deal value, then divide by the cycle length in days: 10 opportunities, a 30% win rate and €9,000 on average over a 90-day cycle give €300 a day.

It is everything that brings in new contacts: SEO, referrals, ads, outbound or directories. It fills the pipeline's inbox, where leads wait until the first call. To be useful, every lead has to carry its source, and each source is judged by the clients it ends up closing, not by the contacts it brings.