Pipeline Velocity: The Formula and How to Use It

Birchline Software has three times its quarterly target sitting in open pipeline and will still miss the quarter. Birchline is invented for this post, and so are all its numbers. None of them is a real result. They show how the arithmetic works.

Pipeline coverage is the ratio of open pipeline value to your target. A common rule of thumb is to keep three to four times your target in open pipeline. The number is easy to calculate, and that is its appeal. The weakness is that coverage counts every open deal equally. A deal that has not moved in two months adds to the total just like one that is about to sign, so a pipeline padded with stale deals looks healthy and misses the quarter anyway.

Pipeline velocity asks a different question. It does not ask how much pipeline you have. It asks how fast that pipeline turns into revenue, and what slows it down.

The formula, with Birchline's numbers#

HubSpot gives the formula ↗ as the number of opportunities, times deal value, times win rate, divided by the length of the sales cycle. The output is revenue per unit of time, so if you measure the cycle in days, you get revenue per day.

velocity = (opportunities × average deal value × win rate) ÷ sales cycle length

Birchline's last quarter, in invented figures: 20 qualified opportunities, an average deal value of 15,000, a win rate of 25% and a 60-day cycle. Velocity is 20 × 15,000 × 0.25 ÷ 60, which is 1,250 per day.

The four levers have different owners.

Do not chase one velocity number. Use it to decide which lever to pull. Plenty of opportunities and a low win rate point to a qualification problem. A high win rate and a long cycle point to a process problem, often late in the deal.

Clean inputs first#

The formula only works if the inputs are honest, and Birchline's founder finds that hers are not. Three fixes come first.

A stage must be a set of facts you can check, such as a named budget owner, a written problem and a booked next step. "We had a call" is not a stage. When stages mean different things to different reps, conversion rates are noise.

Count only qualified opportunities. If anyone with a pulse becomes an opportunity, win rate falls and cycle length stretches, and both are measurement artifacts and not selling problems. To automate that decision, see the lead scoring pipeline.

Then segment before you average. Birchline's blended 60-day cycle hides a 30-day motion for small deals and a 120-day motion for larger ones (again, invented figures). Averages across different motions describe neither.

Where deals die#

Overall win rate says how many deals close. It does not say where the others died. Stage-to-stage conversion does, so Birchline calculates, for each stage, the share of deals that moved on.

Suppose the shape is this. Most leads become qualified opportunities. Roughly half of discovery calls reach a proposal. Only a small share of proposals close. Each pattern points somewhere different:

• A drop between lead and opportunity usually means loose qualification, or marketing and sales using different definitions of a good lead.
• A drop between discovery and proposal often means the buyer has a problem but no budget or reason to act now.
• A drop between proposal and close points at pricing, competition, or stakeholders who never saw the proposal. The proposal-to-signature guide goes deeper on this stage.

Be careful with small samples. With a few deals per stage, one deal moves the percentage a long way. Use a rolling six months, or read the individual deals instead of the rate.

Days in stage#

Average cycle length is the total number of days from creation to closed-won. It is a lagging number, so by the time it rises, you have lost weeks. Days in stage is the early warning. For each open deal, record how long it has sat in its current stage, and compare that with the typical time for won deals in that stage.

A simple rule works for a small team: flag any deal that has spent twice the median days in its stage, and ask one question. What is the next agreed step, and who owns it? If nobody can answer, the deal has stalled, whatever the CRM says.

Stalls cluster. Early ones come from missing urgency or a contact with no authority. Middle ones come from a technical or security review with no owner on your side. Late ones come from procurement, legal or signature, which often start only after you send the paper. Late stalls usually trace back to work you could have started earlier, so ask for the buyer's procurement steps during discovery.

Birchline picks a lever#

With the data cleaned, Birchline's weak point is win rate, and the discovery-to-proposal step is where it leaks. More opportunities would add work without adding revenue, so the founder does not buy a bigger lead list.

We would make the same call. When win rate is low, tightening the target profile beats adding volume, and one clear message across sales and marketing lifts every other input. Messaging alignment starts there.

The change Birchline tests is a written entry rule: three or four facts that must be true before a deal gets a proposal. Typical facts are a named budget owner, a stated cost of the problem, a decision date and an understanding of the buying process. The rule lives in CRM fields, not in a rep's head.

The other levers have their own standard moves.

• Cycle length: build a mutual action plan, a shared list of steps, owners and dates from now to signature. A shared document or spreadsheet is enough. Admin time stretches cycles too, and the sales playbook guide cites Salesforce research on how little of the week reps spend selling.
• Deal value: look at packaging before price. Bundles, tiers and clear scope tend to raise value more than negotiating does. This is a founder decision, so it belongs in positioning work and not in a sales training session.

Four ways the numbers drift#

If a rep recreates an opportunity after a long pause, cycle length looks short and conversion looks high, so keep one record per deal. Renewals close faster and win more often than new business, so report them separately. Deals left open forever inflate coverage and drag down win rate, so close any deal with no activity for a fixed period. And changing stage definitions mid-year breaks comparison with last quarter. If you must change them, note the date and restart the comparison.

None of this needs a tool. It needs a short written definition of each stage and field, agreed once and kept where everyone can find it.

One hour a month#

A spreadsheet is enough. Export closed deals and open opportunities with created date, stage history, source, size and outcome. Calculate the four inputs for the last six months, overall and by segment. List conversion and median days in stage for won deals. Pick the single weakest lever and write one change to test next month.

Then look at the five oldest open deals and decide, for each, to advance it or close it. That step matters most. Closing dead deals lowers your coverage figure and makes your forecast more accurate at the same time.

Change one lever at a time, so you can tell what worked.