Pipeline velocity explained

Pipeline velocity is the speed at which your pipeline produces revenue, expressed in euros per day. You calculate it as the number of open opportunities multiplied by the average deal value multiplied by the win rate, divided by the average sales cycle in days. The figure is useful mainly because it puts the four levers of your sales result into one formula, so you can see which improvement pays off most and where deals get stuck.

What is the pipeline velocity formula?

Pipeline velocity = (number of opportunities x average deal value x win rate) / average sales cycle in days

Each of the four components measures something different:

• Number of opportunities: how many qualified, open deals are in the pipeline.
• Average deal value: what a won deal brings in on average.
• Win rate: the share of deals reaching an outcome that are won.
• Sales cycle: how many days a deal takes on average from creation to win.

Worked example: suppose you have 60 open opportunities, an average deal value of EUR 20,000, a win rate of 25 percent and an average cycle of 90 days. Your velocity is then (60 x 20,000 x 0.25) / 90 = EUR 3,333 per day. Over a 90-day quarter, at this pace your pipeline produces roughly EUR 300,000.

Which lever pays off most

The strength of the formula is that you can compare the levers. In the example above, improving one component at a time by 10 percent:

Mathematically the first three are equal and a shorter cycle is slightly stronger. In practice they differ enormously in what it costs to achieve them.

• More opportunities costs marketing and sales capacity, and often lowers the win rate because the extra opportunities are a poorer fit.
• Higher deal value calls for better pricing discipline or selling more per customer. This is often the cheapest lever, because many businesses give discounts that nobody approved.
• Higher win rate calls for better qualification: not winning more deals, but chasing fewer deals you will not win anyway.
• Shorter cycle calls for less waiting between steps: quoting faster, fewer internal approval rounds, getting the decision-maker to the table sooner.

The value lies in the question you ask next: which of these four is easiest for us to improve by 10 percent?

Where the average misleads you

The formula works with averages, and in B2B averages are treacherous.

Mixed segments. Small deals of EUR 5,000 with a 30-day cycle and large deals of EUR 150,000 with a 200-day cycle together produce an average that fits no real deal. Calculate velocity per segment: new versus existing, small versus large, per product line.

Dead deals in the opportunity count. If half of your open deals are no longer alive, your velocity doubles on paper without a single extra euro coming in. Filter first, as described in how do you spot an unreliable pipeline.

Cycle measured on won deals only. Most people calculate the cycle over won deals. That is correct for the formula, but it hides how much of your time lost deals consume. A lost deal that was open for six months took six months of attention. So also calculate the average time to loss. If it is much longer than the time to win, deals are being given up too late.

Win rate without lost deals. If losses are not recorded, your win rate is artificially high. Treat the formula as optimistic in that case.

Worked example: suppose you split the same pipeline of 60 opportunities into two segments. Expansions with existing customers: 40 opportunities averaging EUR 8,000, win rate 45 percent, cycle 40 days. That gives (40 x 8,000 x 0.45) / 40 = EUR 3,600 per day. New customers: 20 opportunities averaging EUR 44,000, win rate 15 percent, cycle 150 days. That gives (20 x 44,000 x 0.15) / 150 = EUR 880 per day. Together EUR 4,480 per day, and a very different story from the single average: most of your sales result comes from existing customers, while most of the sales conversation is probably about the large new deals. That the two calculations do not add up to the EUR 3,333 from the first example is exactly the point: averages across unequal segments give a figure that fits neither.

Velocity per stage: where deals get stuck

The total figure tells you how fast things move. The breakdown per stage tells you where they stall. Measure per stage:

1. How long deals stay in that stage on average before moving on.
2. What share moves on to the next stage, and what share dies.
3. How that differs between won and lost deals.

You will often see that won deals pass quickly through a particular stage, while lost deals linger there. A deal that has sat in "Quote sent" twice as long as an average won deal has a much smaller chance. That is one of the strongest signals there is, and it is an example of what a model in AI revenue forecasting weighs per deal.

A stage where all deals linger is often an internal bottleneck: a quote that has to pass three departments, a technical validation waiting on one person, a legal review that takes weeks. Those bottlenecks are usually quicker to fix than raising the win rate.

Using velocity for your forecast

Velocity helps with a question the pipeline itself does not answer: how much revenue in this period will come from deals that do not exist yet?

If your cycle is 45 days and your quarter 90 days, part of the quarter's revenue will be won from deals created in the first half of the quarter. Today's open pipeline then underestimates your quarter. Conversely, if your cycle is 180 days, almost all of this quarter's revenue comes from deals that already exist.

That relationship determines how much coverage you need at the start of a period. A short cycle needs less opening pipeline, a long cycle more. How to set that benchmark is covered in pipeline coverage explained.

Velocity is therefore a building block of your forecast, not a replacement. It works with averages and assumes tomorrow will look like yesterday. For how all the building blocks come together, see what is revenue forecasting.

How to get started

1. Choose your period. Calculate the components over the last twelve months, so seasonal effects average out.
2. Filter the pipeline. Only qualified open deals, with dead deals removed.
3. Calculate the four components per segment. At minimum, new customers and existing customers separately.
4. Calculate velocity per segment and in total.
5. Run the 10 percent test from the table above, and next to each lever note what it realistically costs to achieve.
6. Measure time per stage and find the stage where won and lost deals diverge most.
7. Repeat every quarter. Falling velocity often shows up earlier than a missed quarter.

In HubSpot, Salesforce and Pipedrive you can pull most components from standard reports. Time per stage sometimes needs an extra report or an export to Excel.

What velocity does not measure

Velocity stops at won. What happens afterwards falls outside it: whether the deal is invoiced on time, whether a discount is taken off after signing, whether additional work is charged. High velocity with a leak between won and invoiced delivers less than the formula promises. That is another reason why a forecast that ends at invoices is more honest than one that stops at the CRM, a point that also comes up in why sales forecasts are so often wrong.

Frequently asked questions

What is a good pipeline velocity?

There is no general benchmark. The figure depends entirely on your deal size and cycle. Only compare your velocity with your own in earlier periods, per segment.

Should you measure velocity per salesperson?

It can give insight, but be careful. Salespeople with large accounts have a longer cycle and fewer opportunities. Comparing velocity per salesperson without segmenting is comparing apples with pears.

What is the difference from pipeline coverage?

Coverage tells you whether there is enough in the pipeline for your target. Velocity tells you how quickly that pipeline produces revenue. You need both: a lot of pipeline that moves slowly will not make the quarter.

How often should you calculate velocity?

Monthly or quarterly is enough. The figure changes slowly, and weekly fluctuations are mostly noise.