Sales Capacity Planning for B2B SaaS: The Model Most Series B Companies Get Wrong

TL;DR: Most Series B capacity models are built on fantasy math. They assume average quota attainment, ignore ramp curves, and treat attrition as an afterthought. The result is a hiring plan that looks sound in a board deck and falls apart in Q3. Here's how to build one that doesn't.

I've sat in enough Series B board meetings to know what the capacity slide looks like. A clean table. Headcount targets by quarter. A blended quota assumption. A revenue number that adds up perfectly if you squint at it from across the room.

And I've seen what happens when that plan meets reality.

The reps hired in January aren't fully productive until May at the earliest. Two experienced reps resign in February because a competitor recruits them. The new territory carved out of the enterprise segment has no existing pipeline and no brand recognition. By Q3, the revenue number is wrong, the team is stretched, and the CFO is asking why the model was so far off.

It wasn't bad luck. It was bad math.

I offer this view as founder of VEN Studio, former VP of RevOps at a tech unicorn, and a retired seller with seven years carrying my own quota. The capacity planning failures I see aren't random. They follow the same pattern almost every time: leaders confuse quota attainment averages with productive capacity, then build a plan on top of that confusion.

Here's what they're getting wrong and how to fix it.

The Core Mistake: Conflating Attainment with Capacity

Here's the move most Series B revenue leaders make. They take total quota across the team, apply an average attainment rate based on recent performance, and back into a headcount number. If you need $10M in new ARR and your reps carry $800K quotas at average attainment, the math looks straightforward.

The problem is that "average attainment" is a lagging metric describing a historical cohort. It doesn't account for when in the year that attainment happened, which reps drove it, or whether those reps are still on the team. It's a rear-view mirror masquerading as a windshield.

Productive capacity is different. It's the revenue your current and planned headcount can actually generate, given the timing of hires, how long it takes people to ramp, how many people will leave, and whether the territories you're assigning have real demand in them.

Those are four different variables. Most models handle one of them well (quota target), get sloppy on two (ramp and attainment distribution), and ignore the other two entirely (attrition and territory quality).

Variable 1: Ramp Curves Are Not Linear, and Your Model Probably Assumes They Are

Ask most revenue leaders what their ramp period is and they'll say "three months" or "six months." Ask them what productivity looks like during that ramp and they'll go quiet.

Ramp isn't binary. A rep doesn't flip from zero to full productivity on day 91. In most B2B SaaS environments I've worked in, the curve looks something like this:

Month in Role	Realistic Productive Capacity (Illustrative)
Month 1	0% of quota
Month 2	10-15% of quota
Month 3	25-35% of quota
Month 4	50-60% of quota
Month 5	70-80% of quota
Month 6+	90-100% of quota

These are illustrative thresholds, not universal benchmarks. Your curve will depend on deal complexity, sales cycle length, and how good your onboarding is. The point is: this table is not flat, and your capacity model needs to reflect that.

If you're hiring five reps in Q1 and your plan assumes they're fully productive by Q2, you've already overestimated your Q2 and Q3 revenue contribution by a wide margin.

The fix is simple to describe and tedious to build: model each hire individually, assign them a start month, apply a ramp curve to their quota contribution, and aggregate from there. It's more rows in your spreadsheet. It's also the only honest version of the model.

Variable 2: Attrition Is a Planning Input, Not a Post-Mortem Explanation

Voluntary attrition in B2B SaaS sales teams is not a surprise. It happens every year. It happens to everyone. And yet most capacity models treat it as something that happens to other companies.

I've seen Series B plans built with zero attrition assumptions. Not low attrition. Zero. The working assumption is that every rep hired stays hired, ramps fully, and delivers against quota for the entire planning horizon.

That's not a plan. That's a wish.

A realistic capacity model builds attrition in as a structural assumption. You don't need to know who will leave. You need to acknowledge that some percentage will, and that when they do, you lose both their production and the ramp time of their replacement.

The attrition hit is actually double: you lose the departing rep's contribution from the moment they check out mentally (often weeks before they resign) through the end of their notice period. Then you lose 4-6 months of productive capacity while their replacement ramps.

Factor this in. Model a realistic voluntary attrition rate based on your own historical data. If you don't have enough history, be honest about that and use a conservative assumption. Build in a replacement hiring timeline. Assume the replacement ramps like every other new hire.

What you'll find, often, is that your "net new" capacity number is materially lower than the raw headcount math suggested. That's not a problem with the model. That's the model working correctly.

Variable 3: Territory Coverage Is Not the Same as Territory Opportunity

This one is the most underrated variable in capacity planning and the hardest to model cleanly.

Not all territories are equal. A rep assigned to a mature territory with existing pipeline, strong brand recognition, and active renewal accounts operates in a completely different environment from a rep opening a new vertical, a new geography, or a new segment.

The mistake is assigning both reps the same quota and the same attainment assumption. The new territory rep is, in practice, doing a large amount of pipeline generation from scratch. They're not working an existing book of business. Their ramp is functionally longer, their first-year attainment will almost certainly be lower, and your model needs to reflect that.

Before you build your capacity plan, do a territory audit. Ask these questions:

How much existing pipeline exists in each territory?
Is there inbound demand signal (marketing, referrals, brand) for this segment or geography?
Are there active competitive deals already in motion?
What's the average deal size and cycle length in this territory historically?

If a territory scores low on all four, that rep's capacity contribution in year one should be discounted relative to a well-established territory. How much? That depends on your business. But ignoring it entirely is the wrong answer.

Variable 4: Attainment Distribution, Not Attainment Average

The average attainment number hides everything interesting.

I've worked with teams where the "average" attainment looked reasonable on paper, but the distribution underneath it was brutal: one or two reps crushing quota and a long tail of underperformers dragging the mean toward something presentable.

If your capacity model assumes every rep performs at the team average, you're building a plan that only works if the distribution stays perfectly flat. It won't.

Build your model with distribution assumptions instead. Segment your rep population into tiers: strong performers, core performers, and underperformers. Assign realistic quota attainment assumptions to each tier. Then forecast based on how many reps you expect to fall into each bucket.

This is more work than blending to a single attainment rate. It's also dramatically more accurate when something breaks, because it shows you which scenarios create risk and which ones are stable.

A practical starting point, framed as an illustrative example:

Tier	Attainment Assumption	Expected % of Team
Strong	110-130% of quota	20% of reps
Core	75-95% of quota	55% of reps
Underperformer	30-60% of quota	25% of reps

Calibrate those tiers to your own data. But if your model doesn't have tiers, it's not modeling reality.

Building the Model: A Practical Framework

Put this all together and here's what an honest capacity model actually requires:

1. Start with a headcount timeline, not a headcount number. Map each hire to a start month. This matters because a Q1 hire and a Q3 hire don't contribute equally to annual revenue, and your plan needs to show that explicitly.

2. Apply a ramp curve to each hire. Use a curve that reflects your actual sales cycle and your historical ramp data. If you don't have the data, build the curve conservatively, then update it as you accumulate history.

3. Model attrition explicitly. Decide what attrition rate to assume based on your own history. Add replacement hires to the timeline when attrition events are expected. Apply ramp curves to those replacements too.

4. Audit your territories before assigning quotas. Score each territory on pipeline availability, demand signal, and historical performance. Discount first-year attainment expectations for new or underdeveloped territories.

5. Use attainment tiers, not a blended average. Break your rep population into performance tiers. Assign quota attainment ranges to each tier. Forecast revenue using tier distributions, not a single average.

6. Run the model at three scenarios: Base, Conservative, and Downside. Base assumes hiring on schedule, normal attrition, and your expected attainment distribution. Conservative adds a hiring lag of one quarter and bumps attrition slightly. Downside models a meaningful miss on both. Show all three to your board.

What This Actually Looks Like in Practice

At VEN Studio, when we work through capacity planning with Series B teams, the output usually surprises people. Not because the numbers are catastrophically wrong, but because the model surfaces risks that weren't visible before. A hire that was planned for Q2 doesn't hit productive capacity until Q4. An attrition event in a key territory creates a coverage gap that takes two quarters to close. A new vertical that was sized aggressively needs a year, not two quarters, to build enough pipeline to justify the quota assigned to it.

None of these are unfixable. The companies that handle them well are the ones that surfaced them in the model rather than discovering them in the actuals.

The plan doesn't need to be perfect. It needs to be honest.

Frequently Asked Questions

Q: How often should we update our capacity model? At minimum, once per quarter. In practice, any significant hiring event, attrition event, or territory change should trigger a model refresh. A capacity model that's never wrong is a model nobody is updating.

Q: What's a reasonable ramp assumption for a complex enterprise product with long sales cycles? There's no universal answer, but if your average deal cycle is six months or more, expect a rep's first meaningful pipeline contribution to start in month four at the earliest, with quota-level production closer to month nine or ten. Build your plan around that reality, not a three-month ramp assumption borrowed from a mid-market playbook.

Q: Should we assign the same quota to reps in new territories as in established ones? No. New territory reps are doing demand generation and pipeline creation work that established territory reps aren't. Assigning identical quotas without adjusting attainment expectations creates a compensation situation where new territory reps almost always underperform on paper, which accelerates attrition in exactly the roles where retention matters most.

Q: How do we handle the board pressure to show a clean headcount-to-revenue line? Show all three scenarios. Boards that have seen a few planning cycles understand that a single-line plan is a confidence interval of one. Presenting base, conservative, and downside demonstrates rigor, not weakness. It also gives you a defensible answer when Q3 looks different from the original plan.

Q: When does capacity planning start to require dedicated tooling versus a well-built spreadsheet? For most Series B companies with fewer than 30 quota-carrying reps, a well-structured spreadsheet with per-rep rows, monthly time buckets, and scenario toggles is sufficient. The issue usually isn't tooling. It's that the spreadsheet hasn't been built to the level of detail the model actually requires.