What it is
Measured. Where this deal’s pace ranks against comparable deals — same segment, same stage — on a scale of 0 to 100. Read the direction carefully, because it runs the opposite way to every other Beacon score.0 is fastest. 100 is slowest. Lower is better.A deal at 20 has been in its stage fewer days than 80% of comparable deals: it is moving well. A deal at 80 has been there longer than 80% of them: it is slowing, and approaching a stall. The unit is in the name for exactly this reason. A value of 12 is a rank, not a score out of 100 and not a percentage of anything completed.
Why it is published this way
The underlying figure is a percentile of elapsed time, and elapsed time runs the wrong way round: more of it is worse. Beacon could flip it at the point of publishing so that higher meant faster, and matched every other number on the page. It does not. Beacon publishes values as they are defined rather than transforming them on the way out, because a value that is quietly reversed in one place is a value nobody can reconcile with anything else. The direction is stated instead, everywhere the value appears. The practical consequence: do not build an alert on “velocity below 40” by analogy with other scores. Below 40 is fast. Above 75 is the end worth watching.How it is calculated
Measured. The deal’s days in its current stage are placed on the distribution of how long comparable deals took, and the rank read off. The distribution is built from your own deals that closed in the last twelve months — won and lost, both — grouped by segment and stage. It is not built from the deals currently open in your pipeline. Recomputed quarterly. There are no bands on this value. It is a continuous rank. The stall bands you may see beside it belong to days in stage and are built on multiples of the median rather than on percentile position — the two move together but are different scales, and a percentile of 75 does not mean “stalled”. Where fewer than ten comparable deals have closed, there is no distribution of your own. Beacon uses an estimate drawn from your other segments, labels it low confidence, and prevents anything downstream from treating it as calibrated.Where it comes from
No additional source is needed beyond what the benchmark already requires.
How fresh it is
Recalculated on every stage change and again in a full daily pass. The distribution behind it is recomputed quarterly. Every response carries two timestamps.as_of is the moment the value describes. computed_at is when Beacon last worked it out. Read both rather than assuming a cadence.
Currency and rounding
Neither applies. It is a rank on a 0 to 100 scale and holds no money value.What changes it
Time in stage, which pushes the rank up. A stage change, which re-bases the deal against a different distribution and can drop the rank sharply. A change of segment, which moves it to a different distribution entirely. And the quarterly recompute, with nothing happening to the deal. If comparable deals got slower, this deal’s rank falls without it moving any faster. That is the value working correctly — a rank is a statement about a population, and the population changes. What does not change it: deal size, activity, how many people are on the deal, health, or probability.What it is not
- Not a percentage. A deal at 80 is not 80% through its stage and not 80% likely to do anything.
- Not days in stage. That is the count in days. This is the rank that count earns.
- Not the stall band. That is a multiple of the median. This is a percentile. A deal can be high on this and not yet stalled.
- Not a pipeline-level velocity figure. Beacon also measures how fast a whole pipeline is moving, by segment, team or company. That value runs the ordinary way round — higher is healthier — and is not this one. Two velocity numbers, opposite directions, different subjects.
- Not deal probability. Pace feeds probability. A fast deal is not a won deal.