Much of this article is based on a RevOpsAF London session led by Werner Schmidt, CEO and founder of our friends at Lative, alongside a panel of go-to-market strategy and RevOps leaders Ian Matthews of Teradata, Anurag Joshi of Zscaler, and George Erskine of SOPHiA GENETICS. Werner has built sales capacity models at companies like Citrix and Websense, where hundreds of sellers were on the line.
Here's how most sales capacity plans get built:
Someone on the board decides what next year's number should be. That number rolls downhill to the exec team, who add some padding for safety. Then someone divides the padded number by the quota they think a rep can carry, and presto: that's how many sellers you need to hire.
If you’ve been doing this a while, you already know how this ends. Missed quarters, a round of performance plans, and a batch of new hires let go who never had a fair shot.
Then we do it all again next year and act surprised when the same thing happens.
It’s time to stop the cycle and take a more sane approach. While the Top-Down Model (what we just described) isn’t without some value, it must be counter-balanced by a Bottoms-Up approach.
This is a huge point we focused on in our RevOpsAF Podcast Episode with Nancy McBee, CFO at SeekOut. Finance sees the world one, RevOps should see it in another light, and the truth is always somewhere in the middle. The teams must work together to get the numbers right.
Here’s how we recommend budgeting for sales hires and advocating for your plan to your CFO.
The board number isn't the problem. Finance needs a target, the company has a cost base, and somebody has to decide what growth looks like. Like we said above, it’s not a bad start, but it should be pressure tested by a more thorough approach.
Dividing a goal by headcount assumes every rep produces the same amount, in the same amount of time, at the same cost, all year long. None of that is true. Capacity and efficiency vary wildly depending on:
When the math ignores all of this, the gap between plan and reality lands on the sales team. Werner pointed to quota attainment as evidence that capacity planning is failing sales teams. He cited a figure of 48% of sales representatives attaining quota on average. Our goal should be to get it back to 70% or 80%.
That doesn't happen by setting quotas higher. It happens by setting them correctly.
Werner opened his session by asking the room to define sales productivity. The answers he usually gets are activity counts: calls made, emails sent, meetings booked. Or pipeline.
None of these answers are an indicator of productivity.
Activity is motion, and pipeline is a hope with a close date attached.
Productivity is how much revenue a salesperson produces per unit of time. The formula is almost embarrassingly simple:
That's your productivity per FTE.
If you aren't doing this calculation, please start. It's the foundation for every capacity plan, short-range or long-range, because your trailing productivity is the most honest predictor of what your team will produce next.
The catch is that one blended number hides almost everything useful. Sellers work different segments, product lines, regions, and verticals. Productivity needs to be cut by each of those dimensions, because a company that closes huge enterprise deals in two verticals and scraps for mid-market business everywhere else doesn't have one productivity number. It has several, and they tell very different stories about which team the next hire should join.
Productivity tells you what reps produce. Efficiency tells you what that production costs.
Werner told a story from his time at Citrix. His CEO came to him and said he could cut half the sales team and still hit eighty percent of the number. The problem with the assumption was that it wasn’t factoring in the cost of the half he wanted to cut versus their productivity.
Let’s say he was comparing US-based sellers to sellers based in South Africa who work for ⅓ of the cost but produce 40% of the number. Let’s use some simple math to determine who is more productive:
South Africa’s efficiency is 16X while the US has 8X.
After considering the difference in efficiency, the right answer would be to increase the number of sellers overall by reducing headcount in the US and concentrating more hires in South Africa. While it would take more sellers to hit the same number, the cost to the business would make the exchange worth it.
That's a conversation you want to be ready for before sweeping hiring decisions are made by the executive team.
A few practical notes from the session:
Most organizations account for ramp time when they hire new sellers and build compensation plans. That's a good thing. The problem is how they arrive at the number. Too many companies rely on a best guess.
Guessing wrong in either direction is expensive:
The Zscaler panelist described moving away from a fixed ramp assumption toward tenure bands. Reps in months zero to three are modeled with one level of capacity, reps in months three to six with another, and so on. Those bands are combined with historical attainment and attrition rates to produce a capacity number that reflects how people actually ramp, not how we wish they ramped.
You don't need 800 reps to do a version of this. Here's where to start:
When Werner asked the room who factors seasonality into their capacity plans, the response was mostly silence. That's a problem, because finance has a number to hit every quarter, and they don’t always know that what you plan for in Q3 may actually happen in Q4 because entire regions go on holiday during the summer months.
A realistic capacity model accounts for the fact that a selling day in your headquarters’ country isn't a selling day everywhere. At a minimum, build in:
Werner draws a hard line between the two:
He also asked a question I'd love more RevOps teams to ask: why are quotas annual? Werner's goal at Websense, where they scaled to hundreds of sellers very quickly, was the ability to reset quotas every quarter. That takes logic, systems, and a comp plan structure built for it. But quarterly quotas let you tie what reps are asked to carry to current productivity, ramp, and seasonality instead of a guess made eleven months ago.
Most importantly, the order of your calculations matters. Start with productivity and then use it to calculate capacity. Compare capacity to the target. Then you can set quotas. Most organizations run this sequence in reverse, which is how you end up dividing a board number by headcount.
The Teradata panelist described what this looks like in practice. He builds a bottom-up capacity plan using productivity by country and current headcount, then sends it to the CRO and CFO. They send back a top-down target. He tells them it's impossible. Then everyone works out what has to change about productivity to close the gap. It got a laugh from the room, but that's exactly the conversation RevOps should be driving.
If you take one thing away from the article, let it be this: The absolute key to selling your model to the CFO is to show your work. Talk through the model, let them see the math, and give them the data you used to defend it. While other departments may roll their eyes at a data model, your finance team will absolutely appreciate the work and thought that went into a bottom-up approach.
The SOPHiA GENETICS panelist, who previously ran revenue operations at Adobe, made the point that every RevOps leader needs in their back pocket. When a CEO or CFO suggests closing the gap by raising quotas, a good model shows why that won't work. Inflated quotas don't create revenue; they create attrition. Your best sellers leave for a company where they can actually win, and you're further from the number than when you started. He also advocates standardizing quotas where possible, so every seller feels they have an equal opportunity to succeed.
Please, please, please do not start with quotas first. Start with productivity. — Werner Schmidt, CEO and Founder, Lative
This is where your relationship with finance and FP&A becomes non-negotiable. Werner's CFO once handed him fifty new heads a quarter with a very clear (and colorfully worded) instruction not to mess it up. At that scale, a bad capacity model isn't a spreadsheet error. It's a cost base problem finance has to clean up.
Capacity and quotas are only half the math. The other half is whether reps will have enough pipeline to hit what you've asked of them.
The Teradata panelist builds a two-year plan (finance holds a five-year plan) and works backward. He knows what pipeline exists today and roughly what will happen to it: what slips, what gets pushed, what closes, and when. With sales cycles of nine to eighteen months and deals that almost always slip three to six months, he reverse engineers how much pipeline needs to be created each quarter, and by which channel:
He applies the same logic to investment. His models showed that putting $5 million into the India team would return $37 million over three to four years. The same money in the Swiss team wouldn't come close. That's capacity planning connected to ROI, which is the whole point.
The Zscaler panelist described a similar pipeline model. Start with the ACV target, look at current pipeline, apply historical erosion and conversion rates, and the gap tells you how much additional pipeline you need. With solid attribution, that gap turns into marketing's pipeline target and the investment required to hit it.
The SOPHiA GENETICS panelist focuses most of his attention on the qualification stage: the point where an opportunity is accepted and lives in the CRM. If qualification is consistent, conversion rates and velocity become predictable, and so does capacity. That also means marketing-sourced opportunities get held to the same standard as everything else once sales accepts them. No separate rules and no second-class leads.
Here's the uncomfortable part. When Werner asked how many people in the room worked closely with marketing operations, four hands went up. In a room full of revenue operations professionals. RevOps is sales ops, marketing ops, and customer success ops. If marketing ops isn't in your capacity planning conversation, you're modeling a third of the funnel and guessing at the rest.
A former head of RevOps at Oneflow (who now works at Lative) named the other common trap: starting from the system instead of the plan. Teams try to make the CRM spit out answers before they've aligned with finance and the CRO on what the plan actually is. Then the numbers get lost in translation across departments.
Werner was honest that this is hard, even with good tools. There's no silver bullet. But the sequence is clear:
If you're at an earlier-stage company with a dozen reps and eighteen months of history, you won't have enough data for tenure bands or need metro-level efficiency models. Start with the simple productivity ratio, segment it only as rationally required, and revisit it every quarter. A rough model built on real bookings beats a precise-seeming one built on industry averages.
If you're at a larger company where the board number already delivered a goal, and nobody asked for your input, you're not off the hook. Build the bottom-up model anyway. When the plan misses (and it will), you'll be the person in the room with an explanation and a better way to do it next year.
Werner closed with a reminder that stuck with me. Salespeople have families. A cycle of hiring people against unrealistic plans and then firing them when the math doesn't work isn't okay, and it isn't necessary. We have the data to be far more accurate than a board number divided by headcount.
Start with productivity. Understand your capacity. Know what your sellers cost. Then set quotas that give people a real chance to win.
How does your team approach capacity planning? Have you gotten finance to share cost data, or moved to tenure-based ramp? We'd love to hear what's worked for you and where you've hit a wall.
.avif)
.avif)