Sales pipeline

A sales pipeline should tell you which deals are real, what the buyer has done, and what happens next. Most teams use it as a storage bin for every vaguely interested contact.

A 40-person cybersecurity company found this out after hiring its first three account executives. The CRM showed $2.4 million in open pipeline. Half the deals had no meeting scheduled, several accounts were outside the target market, and two proposals had gone to people with no buying authority.

That wasn't $2.4 million in pipeline. It was an unfiltered contact database with optimistic labels.

What a sales pipeline is supposed to show

A sales pipeline is a view of active prospects and opportunities moving through defined stages. Each stage should represent a meaningful change in the buyer's decision process, not just something the rep did.

That distinction matters. "Demo completed" tells you the rep ran a demo. It doesn't tell you whether the buyer cares, who else is involved, or whether anything changed after the call.

A useful stage has entry and exit rules. An opportunity might enter discovery when the buyer agrees to discuss a specific business problem. It should leave discovery when the team understands the current process, the impact, the decision group, and the expected timing.

The pipeline is not the same as a funnel. A funnel shows how a larger audience narrows from awareness to purchase. A pipeline tracks named accounts and the actions required to move specific deals. A forecast is another thing again. It estimates future revenue using pipeline data, conversion history, timing, and judgment.

The CRM only displays the system. It doesn't create one.

The stage mistake teams keep making

Teams usually build stages around internal activity:

• Email sent
• Meeting booked
• Demo completed
• Proposal sent

Those are useful events, but they aren't proof of buyer progress. A prospect can attend a demo out of curiosity, accept a proposal without sharing it internally, then disappear for three weeks.

I think this is where most pipeline advice gets it wrong. It treats activity as evidence. Activity is evidence of effort. Buyer commitment is evidence that a deal may exist.

A better set of stages might look like this in a managed outbound program:

Qualified: The company fits the ideal customer profile, has a relevant problem, and has agreed to discuss it.

Discovery complete: The team knows how the problem works today, what it costs, who is affected, who makes the decision, and why the timing matters.

Evaluation: The buyer has agreed on the criteria for assessing the product and has included the right technical or business stakeholders.

Proposal: Commercial terms have been reviewed with an identified buyer or sponsor. Sending a PDF doesn't qualify.

Negotiation: Pricing, legal, security, procurement, or implementation conditions have named owners and dates.

Closed won or lost: The contract is signed, or the loss reason is recorded clearly enough to be useful later.

The names can change. The evidence can't be vague.

Build the pipeline around a real buying process

Start with how customers actually decide. Don't copy a vendor's seven-stage template and drop it into a six-week mid-market sale or a nine-month enterprise sale.

For example, a 120-person payments company sells reconciliation software to finance teams at online marketplaces. Its average first-year contract is $75,000, the sales cycle is about 90 days, and the quarterly new-business target is $900,000.

The company doesn't need every interested account in the CRM. It needs marketplaces with a transaction-volume problem, a finance owner, and a reason to act now.

One opportunity starts when the team identifies a 300-person marketplace that recently changed payment processors. The finance operations director confirms that reconciliation still happens across spreadsheets and that the processor change created mismatched settlement data. The controller joins the second call. The team learns that finance spends 25 hours each month investigating discrepancies, and that the buying decision is tied to the next audit cycle in four months.

That is a qualified opportunity. Not because someone downloaded a guide. Because the account fits, the problem is real, the impact is understood, and the timing has a reason behind it.

During evaluation, the buyer agrees to review an integration workflow and provide sample transaction data. The technical lead joins. That is stronger evidence than a completed demo because the prospect has committed time and involved another stakeholder.

The proposal includes a $78,000 annual price, implementation milestones, security documentation, and the expected reduction in manual reconciliation work. The controller says procurement must approve the contract. Later, procurement asks for net-60 terms and the security team asks about data retention. Each issue gets an owner and a date.

"Follow up next week" is not a next step. "Controller to review revised terms with procurement on Thursday" is.

That level of detail is what makes a pipeline useful to an SDR team and the account executives taking over later.

What to measure before trusting the forecast

A full dashboard doesn't mean a healthy pipeline. Start with the points where revenue gets stuck.

Stage conversion is the obvious one, but break it down by segment, rep, source, and month when there's enough volume. Conversion from qualification to discovery may look fine overall while enterprise opportunities fail at proposal. That is a different problem from weak top-of-funnel sourcing.

Time in stage is useful too. Use the median where possible. One stalled enterprise deal can make an average look ridiculous. Track slippage separately. If deals keep moving from March to April to May, the problem is usually weak qualification, a missing decision-maker, or no control over the next step.

Pipeline coverage is open qualified pipeline divided by the target. A team with a $300,000 quarterly target and $900,000 in genuinely qualified opportunities has 3x coverage. A team with $900,000 in old records and no buyer activity doesn't.

Stale opportunity rate deserves attention. Set a rule, such as no buyer activity or scheduled next action for 21 days, then review those deals rather than letting them sit open forever. Win rate should be measured by opportunity count and revenue. Winning ten small deals can hide a serious miss against the revenue target.

Forecasting comes after this cleanup. A forecast based on every open CRM record is fiction with decimal places.

Pipeline reviews that don't waste everyone's time

Run a weekly review, but don't read every record aloud. Focus on deals with a meaningful change, a late-stage risk, no next action, or an unusual amount of time in stage.

Ask what the buyer did since the last review. Then ask what the buyer must do next, who else is involved, and what could stop the deal from closing on the stated date. If the rep can't answer, the opportunity isn't forecast-ready.

Keep one source of truth. A spreadsheet can work for a five-person team. Sales pipeline templates in Excel can help define fields such as account, stage, value, close date, owner, and next action. But once several reps are editing records, the file becomes hard to audit and easy to duplicate. It also won't show activity history reliably.

Use a CRM or a simple shared system that the team will actually update. The fields should support decisions, not collect trivia. TechTarget's sales coverage of pipeline stages makes the same practical point: stages need to match business requirements and the real workflow.

Loss reasons should be specific. "No decision" tells you almost nothing. "Budget removed," "incumbent retained," "security failed," "no executive sponsor," "timing moved," and "poor fit" point to different fixes.

For outbound teams, record the source and message angle too. An account created through cold outreach after a funding round should not be treated like an inbound demo request. Those opportunities may have different intent, sales cycles, and conversion rates. That difference belongs in the pipeline, not in someone's memory.