How to Build a Sales Pipeline That Shows the Truth

What is a sales pipeline, and what should it tell you?

A sales pipeline is a stage-by-stage view of every open deal, with its value, owner, next step and expected close date. Salesforce's training material describes the record behind it, the opportunity, as a qualified lead that has progressed through the first stages of evaluation, with expected revenue, a sales stage, a close date and a probability of closing.

A good pipeline lets an owner answer three questions in a few minutes: how much can we realistically close this quarter, which deals need help this week, and where do deals tend to get stuck? If the answers take a round of calls to the sales team, the pipeline is not doing its job.

Two things sit outside it. Unqualified enquiries belong on your lead list, because mixing them in inflates the total, and our guide to managing and qualifying leads covers everything up to the moment a lead becomes a deal. Choosing software is a separate decision too, covered in how to choose a CRM. The method below works in a spreadsheet as well as in a CRM, though a spreadsheet gets harder to keep honest once several people edit it.

How do you choose stages based on what the buyer has done?

Name each stage after something the buyer has done, because a sale moves forward only when the buyer does. A stage called Quote sent becomes true the moment your rep presses send, even if nobody opens the file. A stage called Proposal reviewed means the buyer read it and came back with a reaction, and that tells you far more about the deal.

To find your stages, take your last ten or so won deals and write down what the buyer did before each one closed: who joined the calls, when finance got involved, when the price discussion began. Steps that appear in most of them are your candidates. If an owner or finance manager has to approve the purchase but never attends the first meeting, give that step its own stage so it cannot stay hidden.

Here is a starting set for a business that sells to other companies. Change the wording to match how your buyers behave:

Need confirmed: the buyer has described the problem in their own words and agreed to another conversation.
Decision maker engaged: the person who approves the purchase has joined a call or meeting, or has replied directly.
Proposal reviewed: the buyer has read your proposal and responded with questions, changes or a clear reaction.
Terms in negotiation: the buyer has raised specific points on price, scope, payment or the contract.
Verbal commitment: the decision maker has said yes, and only paperwork or internal approval remains.
Won or Lost: the contract, purchase order or first payment has arrived, or the deal has ended with a recorded reason.

How many stages and pipelines do you need?

Use as few stages as your buyers' real steps allow, and add one only when reviews keep exposing a hidden step. HubSpot's default sales pipeline has seven stages, counting Closed won and Closed lost. No rule says yours must match it, but it is a fair ceiling to begin with, and that ceiling is our judgment. Remove any stage that deals regularly skip.

Run a separate pipeline only when the buyer's steps truly differ, such as new business and renewals, or government tenders and private sales. Forcing a tender into a stage list built for private sales leaves deals sitting in stages that do not describe them.

What exit criteria should each stage have?

An exit criterion is a fact that must be true, and visible on the record, before a deal can move on. Write each one so that two managers reading the same deal would give the same answer. "Buyer seems interested" fails that test. "Buyer named the person who signs" passes it.

These match the stages above. They are a sample to adapt, and each should be something a rep can point to, such as a meeting note, an email or a WhatsApp message:

Leaving Need confirmed: the problem is on the record in the buyer's words, the people involved in the decision are named, and the next meeting has a date.
Leaving Decision maker engaged: the approver has confirmed the need and when they want it solved, and has agreed to receive a proposal.
Leaving Proposal reviewed: the buyer has responded to a specific proposal, the amount on the record matches it, and the close date comes from the buyer's own timeline.
Leaving Terms in negotiation: open points are settled, and the buyer has named any remaining internal steps, such as finance or legal review, with dates.
Leaving Verbal commitment: a signed contract, purchase order or first payment is on file, and the deal closes as Won.
Closing as Lost: a reason is chosen from a fixed list, with a short note on what happened.

How do you keep exit criteria honest?

Move a deal backward when the evidence goes away, and limit what you force people to fill in. If your contact leaves and the new manager has agreed to nothing, that deal no longer belongs in Terms in negotiation, whatever it looked like last month.

Software can help with the second part. HubSpot lets you set required properties that appear when a deal moves to a stage, and users cannot save the record until they enter a value. That works well for one or two facts per stage. Ask for ten and reps will start typing anything to get past the form, which leaves you with fields you cannot trust.

How do you set stage probabilities?

A stage probability is the share of deals reaching that stage that you went on to win. If 40 deals reached Proposal reviewed last year and 16 were won, that stage's probability is 40%. A HubSpot article makes the same case for using a stage's conversion rate as its win probability, on the condition that the pipeline is well designed and the rate stays fairly stable.

Until you have that history you are estimating, and you should say so. Vendors' sample settings show how much starting points differ: HubSpot's default gives Appointment scheduled 20%, while Salesforce's sample mapping runs from 10% for Prospecting to 90% for Negotiation/Review. Use figures like these only as placeholders, then recount every quarter or so. Be careful with small numbers: with ten deals, one win moves the rate by ten percentage points.

Tie the probability to the stage and keep reps from changing it deal by deal. Salesforce changes probability and forecast category when the stage changes. HubSpot also updates deal probability with the stage, until someone edits it by hand and it stops following the open stages. A rep who feels sure about a deal should show it by meeting the next exit criterion.

How do you turn the pipeline into a forecast?

Multiply each deal's amount by its stage probability and add up the results, which HubSpot calls the weighted amount. Count only deals expected to close in the period you are forecasting, and convert everything to one currency first if you sell in SAR, JOD, IQD or USD.

This example is made up. The company, deals, percentages and target are invented to show the arithmetic, and they are not benchmarks. Imagine a firm in Jeddah with 30 open deals due to close this quarter:

Need confirmed: 12 deals worth SAR 600,000, at 10%, gives SAR 60,000.
Decision maker engaged: 8 deals worth SAR 480,000, at 25%, gives SAR 120,000.
Proposal reviewed: 5 deals worth SAR 350,000, at 40%, gives SAR 140,000.
Terms in negotiation: 3 deals worth SAR 240,000, at 70%, gives SAR 168,000.
Verbal commitment: 2 deals worth SAR 200,000, at 90%, gives SAR 180,000.

How do you read the numbers in that forecast?

Read the weighted total as an estimate across many deals, and read the committed total as what you can rely on. In the made-up example the full pipeline comes to SAR 1,870,000 and the weighted forecast to SAR 668,000. Nobody should plan on the first number. The second gets less reliable as the deal count falls, since any single deal closes in full or not at all.

Forecast categories give you the second view. Microsoft's Dynamics 365 documentation uses Pipeline for early or stalled deals, Best case for deals that are progressing with no commitment yet, and Committed for deals where the customer has made a verbal or contractual commitment. Applied to the example, Committed is the SAR 200,000 in Verbal commitment, and Best case is the SAR 590,000 in Proposal reviewed and Terms in negotiation.

Now compare with the target, as Microsoft suggests doing with Committed and quota. Against an invented target of SAR 700,000, committed deals leave a gap of SAR 500,000. Closing it from Best case alone would need about 85% of that value to close, while this company's own probabilities for those stages are 40% and 70%. The manager can say early that the quarter is at risk and decide what to do about it.

How do you run a weekly pipeline review?

Meet at the same time each week for about 30 minutes, work from the live pipeline, and spend the time on what changed and what is at risk. Reps update their deals before the meeting, so nobody reads records aloud. Salesforce's Pipeline Inspection training shows the habit worth copying: compare the pipeline with an earlier date and look at what moved in and out of the period, including deals pushed out of the quarter. Without a tool, save last week's totals in a sheet and compare by hand.

A simple agenda:

Start with weighted and committed totals against target, compared with last week, and ask what caused each change.
Go through Committed deals first, as Microsoft suggests, and look for anything that has stalled.
Check deals whose close date has moved or passed. HubSpot's article says an open deal with a past close date means something is clearly wrong, and it defines a pushed deal as one the team expected last month and now expects this month.
List open deals with no dated next step. Each one leaves the meeting with a next step or gets closed.
Pick two or three large deals and ask what the buyer has done since last week, answering from the record.

What makes weekly reviews keep working?

Reviews keep working when reps can report bad news safely. If moving a deal backward or losing it brings a telling-off, reps will leave it open, and the forecast pays for that. Thank people for closing a dead deal early, because it frees their week for deals that can still be won.

Record each decision on the deal itself, so it is there next week. Once a month, step back and look at how long deals stay in each stage and at your loss reasons. The stage where deals pile up shows you where to coach, or where your offer needs work. Microsoft's guidance on lost revenue makes the same point: analyze the patterns for coaching and pipeline strategy.

How do you clean out stale deals?

A deal is stale when the buyer has done nothing for longer than deals normally spend in that stage, so set the limit from your own numbers. HubSpot's article suggests using average time in stage and the average age of won deals to set stale-deal triggers. A simple way to start is to flag any deal that has spent twice the average time in its stage. That multiplier is our suggestion, so adjust it to your sales cycle.

Watch what counts as activity. Pipedrive's rotting feature flags deals idle for a set number of days per stage, and actions such as completing an activity or adding a note reset the count. HubSpot's last activity date also includes logged notes, calls and meetings. So a rep's note can make a dead deal look alive, and your check should also ask when the buyer last responded.

Give each flagged deal one of four outcomes within a week:

Revive it: the rep contacts the buyer, the buyer responds with something concrete, and a dated next step goes on the record.
Move it back: the buyer is still interested, but the evidence for the current stage has gone, so it returns to the stage it really fits.
Park it: the buyer has given a real reason and a date, such as next year's budget. Microsoft's Omitted category exists for deals on hold, keeping them out of forecast totals. Set a date to return to it.
Close it as lost: after the attempts you agreed on, such as a call and a final message over two weeks, close it with a reason like no budget, chose another supplier, no decision or stopped replying.

What happens after you clean the pipeline?

The total falls, so warn the owner before the first big clean-up. No sales have disappeared. Those deals were never going to close this quarter, and the forecast now says so.

Keep the records and avoid deleting them. Microsoft's documentation lets you reopen a closed opportunity if the customer becomes interested again, and HubSpot's closed lost reason field can hold your own list of options. Those reasons become useful data within a few months. Repeat the clean-up monthly, and it will stay a small job.

How does Repzo handle this?

The sales CRM in Repzo Workstation lets you build custom stages in Kanban and list views, and separate pipelines for new business, renewals or regions. Each deal holds its amount, probability, forecast category, expected close date and next step, with an owner and reminders, and deal values can be in multiple currencies. Activity history is linked across accounts and contacts, so a review can check what happened on the deal.

For reviews, the feature page describes weekly pipeline reviews without exporting spreadsheets, win and loss analytics, stage velocity and spotting stalled opportunities. Analytics and reporting covers deals, with filters such as date range, owner and pipeline, and groupings by week, month and quarter. A qualified lead converts in one click into a linked account, contact and deal.

You can test your own stages during the 7-day free trial. Our public pages mention standardizing deal stages and required handoff data, so if you want a stage to block a move until its exit criteria are filled in, or want to know how stale deals would get flagged, ask our sales team to show you how that would work with your stages.

Frequently asked questions

How many stages should a sales pipeline have?

No fixed number exists. Many businesses can start with about five open stages plus Won and Lost, built from what buyers actually did in recent wins. Add a stage only when reviews keep showing a hidden step, and remove any stage deals regularly skip.

What is the difference between a sales pipeline and a sales forecast?

The pipeline is the list of open deals by stage. The forecast is your estimate of what will close in a period, calculated from those deals. Stages that do not reflect buyer actions will give you a poor forecast whatever formula you use.

Should reps set the probability on each deal themselves?

Usually not. Tie the probability to the stage so the same evidence produces the same number across the team. A rep can still add judgment by choosing a forecast category or writing a note, which keeps their view visible without changing the math.

Can a deal move backward in the pipeline?

Yes, and it should when the evidence for its stage goes away, for example when your contact leaves or the buyer reopens the scope. Moving it back keeps the forecast accurate and shows the rep what has to happen next.

How long before a deal counts as stale?

It depends on your sales cycle, so use your own average time in each stage. A two-week silence may be serious for a small repeat order and normal for a large contract waiting on a board meeting. Set a separate limit for each stage.

Can we run a pipeline in a spreadsheet?

Yes, for a few reps and a modest number of deals. Use one row per deal with stage, amount, owner, next step and date, and expected close date. It gets harder when you need reminders, activity history, or a view of what changed since last week.

How to build a sales pipeline that shows the truth about your deals

The short answer

Key takeaways