Derived from the model, not reverse engineered to a number.
The most common method is also the least defensible: take the number the board wants, divide by the reps you have, add a stretch. It produces a quota nobody believes and a plan with no model underneath it.
When it is challenged, and it will be, there is nothing to point at.
Ramp time, by segment. Historical win rate, by segment and source. Territory potential from the market model. Attrition, from what actually happened rather than what you hope.
Those four produce a capacity number. The quota is derived from it. If the derived number is below the board number, that gap is the real conversation and it is better to have it in October than in May.
A plan that only works if nothing changes is not a plan. Show what happens if hiring slips a quarter, if a top rep leaves in March, if win rate drops two points.
Putting those on the page is what turns a quota into something a CFO will defend alongside you.
The last step is checking that the comp plan pays for the behaviour the operating plan wants. If the plan is expansion and comp pays flat on new logo, the plan loses.
This is usually where the real misalignment sits, and it is almost never caught during quota setting.
From a capacity model built on ramp time, win rate by segment, territory potential and attrition. The quota is derived from what the team can actually produce, not from the target divided by headcount.
Three to five times is commonly quoted in B2B SaaS, but the ratio is an output of the model rather than an input. If your derived capacity produces a different ratio, the ratio is telling you something.
That gap is the most useful output of the exercise. It is either a hiring plan, a pricing change, or a conversation about the target. Finding it in planning is much cheaper than finding it in Q2.
Model each rep individually by start date and segment ramp curve rather than applying an average. Averages hide the quarter where three new hires are all still ramping at once.
Rarely, and only with a written reason. Changing it to chase a number destroys trust in the model. Changing it because the territory materially changed is defensible.
We use cookies to run the site and to understand how it is used. You can accept all or necessary only. See our privacy notice.