How to Build a Sales Pipeline in Your CRM

Most CRM pipelines start out tidy and turn into fiction within a few months. Deals sit in “Proposal” for 200 days. Half the close dates are in the past. The forecast says you’ll have a record quarter, and then you don’t.

The problem usually isn’t the CRM. It’s that the stages were never defined well enough for 2 people to agree on where a deal belongs. Fix that, and the pipeline becomes a tool you can actually run the business on.

This guide walks you through building a sales pipeline from scratch: how to define stages, write exit criteria, choose required fields and set up the reports that keep it accurate.

What a sales pipeline is (and isn’t)

A sales pipeline is a visual view of every open deal, organized by where each one sits in your sales process. In a CRM, it usually appears as columns (stages) with deal cards you move from left to right.

It does 3 jobs:

• Shows what to do next. Every deal in a stage has an obvious next action.
• Forecasts revenue. Deal values and stage probabilities add up to an expected number.
• Reveals where deals get stuck. Conversion between stages shows your weakest step.

A pipeline is not the same as a sales funnel. A funnel describes how buyers move from awareness to purchase, mostly from the marketing side. A pipeline tracks specific deals your team is actively working. If you’re new to CRMs in general, start with our guide on what a CRM is.

Step 1: Map how your buyers actually buy

Before touching the CRM, grab your last 10 to 20 won deals and write down what happened in each, in order. Who reached out first? When did you send pricing? Was there a site visit, a demo or a trial? Who signed?

You’ll usually see 5 to 7 milestones that repeat. Those are your stages. Base them on what the buyer does, not on what your team does internally. “We sent a proposal” tells you about your activity. “The buyer reviewed the proposal with the decision-maker” tells you about the deal.

If you sell very different things (for example, one-off installs and annual service contracts), you may need separate pipelines. Otherwise, keep one.

Step 2: Define stages with exit criteria

This is the step most teams skip, and it’s the one that matters most. Each stage needs an exit criterion: a specific, checkable event that must happen before a deal moves forward.

Here is a sample pipeline for a small B2B or high-ticket service business. Adjust the names and criteria to fit how you sell.

The probabilities above are placeholders, not benchmarks. Once you have a few months of data, replace them with your own historical rates: of all deals that reached a stage, what share eventually closed?

Key takeaway: If a deal can move to the next stage without the buyer doing anything, your exit criterion is too weak. Good criteria describe a buyer action you could prove in a CRM note.

Step 3: Set up the pipeline in your CRM

The steps are similar across tools. In HubSpot, for example, you edit deal stages in the pipeline settings, set a deal probability for each stage and keep both a won and a lost stage so reports handle closed deals correctly. Salesforce works the same way in principle: each opportunity stage has a default probability, and expected revenue is calculated from the deal amount and that probability.

A practical setup checklist:

1. Create or rename stages to match your table from Step 2.
2. Add the exit criteria to each stage’s description, or pin them in a shared doc your team can see.
3. Set stage probabilities. Start with estimates and update quarterly from real data.
4. Make key fields required at specific stages. For example, require deal amount and close date before “Proposal sent,” and a loss reason before “Closed lost.” Many CRMs let you require properties when a deal enters a stage.
5. Create a loss reason list. Use a dropdown (price, timing, chose competitor, no decision, not a fit) rather than free text so you can report on it.
6. Automate the busywork. Create a follow-up task when a deal enters a stage, or alert the owner when a deal hasn’t moved in a set number of days. See what marketing automation can do for more ideas.
7. Set permissions. Decide who can edit stages and probabilities. Usually just 1 admin.

If you’re still choosing a system, our guides on how to choose a CRM and HubSpot vs Salesforce cover pipeline features worth comparing.

Step 4: Decide which fields every deal needs

Too many required fields and your team stops updating the CRM. Too few and your reports are useless. These 6 cover most small teams:

• Deal amount: your best estimate of the contract value.
• Expected close date: a real date, updated when it slips.
• Deal owner: 1 person accountable.
• Next step and date: what happens next and when.
• Lead source: where the deal came from, so you can judge marketing channels.
• Loss reason: required only when closing a deal as lost.

A “next step” field is the single best habit for pipeline accuracy. A deal with no next step on the calendar isn’t really active.

Step 5: Build the 5 reports that keep it accurate

Reports turn the pipeline from a to-do list into a management tool. Set up these 5 and review them weekly.

A quick note on coverage. If your win rate from “Qualified” is around 25%, you need roughly 4 times your target in qualified pipeline to hit it. That’s simple math: target divided by win rate. Use your own win rate rather than a generic rule.

If you want these numbers on 1 screen, see our guide to building a business dashboard.

Step 6: Run a weekly pipeline review

A 30-minute weekly review keeps data clean and deals moving. Use the same agenda each time:

1. Stalled deals first. For each, decide: next step booked, or close as lost.
2. Past-due close dates. Update every one.
3. Deals moving stages this week. Confirm the exit criterion was really met.
4. Biggest deals. What could block them, and who’s helping?
5. Forecast check. Does the weighted pipeline still support your target?

Closing dead deals is healthy. A smaller, honest pipeline beats a large, imaginary one, because forecasts and staffing decisions depend on it.

Common pipeline mistakes

• Too many stages. If 2 stages always happen together, merge them.
• Activity-based stages. “Called twice” is an activity, not a stage.
• No closed lost discipline. Deals that never close should be marked lost with a reason, not left open forever.
• Optimistic probabilities. Without real data, people round up. Recalculate from history every quarter.
• One pipeline for everything. Renewals, new business and partner deals often move differently. Separate them if the stages genuinely differ.

Where to start this week

1. Pull your last 10 to 20 won deals and list the milestones.
2. Draft 5 to 7 stages with exit criteria in a shared doc.
3. Get your team to agree on the definitions before you change the CRM.
4. Update the CRM stages, required fields and loss reasons.
5. Clean up existing deals: re-stage, re-date or close.
6. Book a recurring 30-minute pipeline review.

Pricing for pipeline features varies by CRM plan, and extra pipelines or custom reports are often limited to paid tiers. Our CRM cost guide explains the pricing models.

Conclusion

A useful sales pipeline comes down to definitions. Build stages around what buyers do, attach a clear exit criterion to each, require only the fields you’ll actually use and review the same 5 reports every week. Do that, and your CRM stops being a guessing game and starts telling you which deals to chase, which to drop and how much revenue is really coming.

Sources

• HubSpot Knowledge Base: Set up and customize pipelines — deal stages, deal probability and won/lost stage requirements
• Salesforce Help: Opportunity fields — stage probability and expected revenue definitions
• Salesforce Trailhead: Create and manage stages and sales processes — default stage probabilities and sales processes