Pipeline velocity: the math of sales

Two salespeople in the same team have the same 30% revenue drop. One works harder — more calls, longer days. It does not help.

The other checks four numbers and sees that 40% fewer conversations entered her pipeline six weeks ago. Everything else stayed the same. She fixes that specific input and revenue recovers.

The difference is not talent — it is knowing which part broke. This lesson introduces those four parts and the formula that connects them.

Revenue is not one number that goes up or down — it is the product of four separate inputs, each of which changes independently.

Opportunities (volume). How many prospects are actively in conversation with you right now. A pipeline with 50 conversations will produce more closes than a pipeline with 10, assuming the quality is similar — because each conversation is a chance, and more chances mean more outcomes.

Deal size (value). The typical revenue from a single closed deal. Many of the steps in a sale — discovery, proposal, negotiation — take roughly the same amount of time whether the deal is $5,000 or $50,000. Larger deals improve the return on each hour spent.

Win rate (conversion). The percentage of pipeline opportunities that end in a closed deal. If 100 prospects enter and 10 buy, the win rate is 10%. A drop from 10% to 5% halves revenue even if everything else stays constant.

Cycle length (speed). How many days from first contact to a signed deal. Shorter cycles mean revenue arrives sooner — and fewer deals stall in the middle, where most of them quietly die.

These four inputs combine into a single formula:

Pipeline Velocity = (Opportunities × Deal Size × Win Rate) ÷ Cycle Length

Example: 50 opportunities × $10,000 × 20% win rate ÷ 30 days = $3,333 per day.

Because the inputs multiply rather than add, small improvements across all four compound faster than expected:

• Improve each lever by 20%
• Opportunities: 50 → 60
• Deal size: $10,000 → $12,000
• Win rate: 20% → 24%
• Cycle length stays at 30 days
• Old velocity: $3,333/day → New velocity: $6,912/day

The total roughly doubles — not because any single input changed dramatically, but because 1.2 × 1.2 × 1.2 × 1.2 ≈ 2.07. You do not need to double your hours to double your revenue — you need to slightly turn all four dials.

But this math assumes every opportunity in the pipeline is real. If the data going into the formula is fiction, the output is fiction too.

The formula above treats every opportunity equally. If the pipeline says 100 deals, it calculates as if all 100 can close. In practice, they cannot.

A prospect who says "sounds interesting" but has no budget is not an opportunity — it is a conversation that feels productive but has no mechanism to turn into money. A seller with 100 such conversations is busier than a seller with 30 qualified ones, but the second seller closes more deals — because every hour goes toward prospects who can actually buy.

The formula itself reveals the problem. If the pipeline shows 100 deals at a 10% win rate, the prediction is 10 closes. But if 60 of those deals lack budget or authority, the real pool is 40 — and the actual result will be closer to 4.

The Opportunities variable was inflated, and the Win Rate drops because unqualified deals almost never convert. The formula looked healthy; the data behind it was not.

The early filter

Four conditions separate a real opportunity from a dead end:

• Urgency: A problem without a deadline gets postponed indefinitely — no urgency means the deal will stall, no matter how good the pitch.
• Authority: If the contact cannot approve the purchase, the deal depends on someone the seller may never reach or influence.
• Budget: Interest without allocated money is a conversation, not a deal — no amount of selling creates a budget that does not exist.
• Consequence: If the prospect can live with the problem comfortably, the cost of changing will always feel higher than the cost of doing nothing.

Any missing condition is a signal. A prospect who has all four is worth ten who have only one — because the seller's time is the one resource that cannot scale.

Qualification protects the Opportunities and Win Rate variables. But even with qualified deals, the formula has a third vulnerability: the Cycle Length variable is only accurate if pipeline stages reflect reality.

Cycle length is calculated from when a deal enters a stage to when it exits. If stages are subjective, two sellers log the same moment differently — and the cycle length data becomes noise.

"Interested" is not a stage. It is one seller's interpretation of a polite email reply. Another seller might call the same reply "engaged." If both log their deals as "Stage 2," the pipeline data mixes two completely different situations — and the forecasts built on that data are unreliable.

Stages work when they describe events that either happened or did not:

The left column is a story the seller tells. The right column is a fact anyone can verify. When the whole team uses fact-based stages, two things happen: forecasts become accurate, and deals that are stuck become visible — because a deal sitting at "discovery completed" for three weeks is obviously not progressing.

This fiction is compounded by ghost deals. Prospects stop responding, choose a competitor, or simply lose interest — but nobody updates the stage. Ghost deals inflate the Opportunities count and stretch the Cycle Length, making the formula optimistic on both ends.

A practical rule: if a prospect has not responded in 14 days despite follow-up, they are no longer an active deal. It is not dead forever — but counting it as alive corrupts every forecast.

When qualification is strict and stages are fact-based, the formula becomes a reliable diagnostic tool. But like any tool, it has limits.

New products without history. If no deals have closed yet, there is no data to feed the formula. Early-stage sales is experimentation — testing audiences, messages, price points — not optimizing a formula that has no inputs yet.

Complex enterprise deals. Sales cycles of 12 months with committees, procurement processes, and internal politics do not reduce to multiplication. The formula can track these deals, but relationships and timing matter more than any single lever.

Over-optimization. When the metrics become the goal instead of the tool, the system works against itself. A seller who optimizes response speed but has nothing useful to say in those fast responses is gaming the number. The formula is diagnostic — it shows where to look, not what to do once you find the problem.

The formula reveals which part of the process is underperforming. But one lever — win rate — depends on something no formula captures: who controls the conversation. That is the subject of the next lesson.