How do you do sales capacity planning for the next fiscal year in 2027?

What capacity planning actually is and why it decides the fiscal year

Capacity planning is the arithmetic bridge between a revenue number someone else committed to and a headcount number someone else has to fund. Finance sets a bookings target. RevOps translates it into a specific count of quota-carrying humans, sitting in specific segments, starting on specific dates, producing at a specific fraction of quota. Everything downstream — territory design, quota assignment, comp plan cost, pipeline coverage ratios, recruiting requisitions, onboarding capacity, sales enablement calendars, even office and tooling spend — inherits from this one model. Get it wrong and you spend the entire fiscal year managing symptoms of a planning error nobody remembers making.

The reason it goes wrong so consistently is that the model has a comfortable answer and an honest one, and the comfortable answer is cheaper. Take a $30M new-bookings target with $1.2M quotas. Assume everyone hits 90% and everyone is fully ramped: you need about 28 reps. Now use a realistic 65% attainment and a 0.85 blended ramp factor: $30M ÷ ($1.2M × 0.65 × 0.85) = roughly 45 reps. That is a 17-head difference on the same revenue target — call it $4M-$6M of fully loaded cost including comp, benefits, tooling, and management overhead. The gap between those two numbers is not a rounding error. It is the difference between a sales org that makes plan and one that spends Q3 explaining a shortfall.

The word "capacity" is doing specific work here. It does not mean headcount. It means *productive selling capability available during the fiscal period*. A rep hired in October has a headcount of one and a capacity of roughly zero for the fiscal year they were hired into. A rep on a PIP has a headcount of one and a capacity somewhere between zero and half. A rep covering two territories after a departure has a headcount of one and a capacity that is not two. Capacity is a time-weighted, productivity-weighted number, and the moment you let it collapse into a simple headcount count, the model stops describing reality.

There is an adjacent version of this problem worth noting, because the same math governs it: SDR capacity, solutions-engineering capacity, customer-success capacity, and professional-services capacity all follow the identical structure — demand ÷ (throughput per person × realization rate × ramp factor). If your AEs need 3.5x pipeline coverage and your SDRs generate 12 qualified opportunities per month at 70% acceptance, the SDR count falls out of the AE count mechanically. Teams that model AE capacity in isolation routinely under-build the support functions and then wonder why fully staffed AEs have empty calendars. Plan the whole revenue engine, not one seat type.

The strategic value of getting this right extends past making the number. A defensible capacity model is the single most useful artifact RevOps produces in the annual planning cycle, because it converts an argument about ambition into an argument about assumptions. When the CFO wants fewer heads, the conversation stops being "sales always asks for more people" and becomes "which input do you want to change — the target, the quota, the attainment assumption, or the ramp?" Those are answerable questions. "Do we need 45 or 28 reps?" is not.

The step-by-step process from close of Q3 to requisitions open

Step one — lock the demand side. Get the board-approved new-bookings target segmented by enterprise, mid-market, and SMB. Do not accept a single blended ARR number; enterprise and SMB have entirely different capacity math and blending them guarantees a mis-shaped org. Separate new logo from expansion, because expansion capacity often lives in a different team with different productivity. If finance hasn't segmented it yet, propose a split based on last year's actual mix and make them argue with it — that is faster than waiting.

Step two — pull the honest productivity baseline. Export trailing-twelve-month attainment from the CRM at the individual rep level, not the team roll-up. You want the distribution, not the average: top quartile, middle two quartiles, bottom quartile. In most B2B SaaS orgs the top 20% of reps produce 50-60% of bookings and the bottom half lands at 40% of quota or below. Planning with the mean silently assumes every future hire will be median or better, which no hiring process on earth delivers. Exclude reps with under six months of tenure from the baseline or you double-count ramp drag later.

Step three — build the ramp curve from your own cohort data. Take your last three hire classes and chart time-to-first-deal and time-to-full-quota. The typical pattern is near-zero contribution for the first 90 days, then 6-9 months to full productivity, netting out to 60-75% of quota delivered across the first twelve months. Convert this into a monthly ramp table rather than a single fudge factor — a February start and an August start have wildly different fiscal-year contributions, and a single blended multiplier hides that.

Step four — set the attrition assumption from HRIS data, split voluntary and involuntary. Annual AE attrition in B2B SaaS structurally runs 25-45%. Involuntary attrition is partly a planning input you control (how fast you cut underperformers); voluntary attrition is a market condition you mostly do not. Model them separately because the timing differs — involuntary clusters after Q1 and Q2 quota reviews, voluntary clusters after commission payouts and around Q1 and Q3.

Step five — compute base headcount by segment, then apply the buffer. Run the core formula per segment, sum, then add 25-30% on top for attrition. Time the incremental hires against predicted departure months rather than loading them all into January. Yes, some quarters will look over-resourced on paper. That is the point — the alternative is entering Q3 with a hole recruiting physically cannot close.

Step six — reconcile against the constraints that will actually bind. Can recruiting source that many qualified candidates in the window? Can enablement onboard 15 people in one month without the training quality collapsing? Are there enough accounts and enough territory to give each new rep a viable book? A capacity plan that requires hiring 40 enterprise AEs in a market where you historically close 6 hires a quarter is fiction with a spreadsheet attached. Cap the plan at what the funnel can actually deliver and escalate the delta explicitly.

Step seven — pressure-test, then lock. Run the scenarios (covered below), present the base case with two stress cases, get the CFO's sign-off on the assumptions rather than just the number, and open requisitions. Then instrument the model so you can measure drift against it every month.

Costs, timelines, and the ranges that show up in real plans

The timeline is the part most teams underestimate, and it is the reason capacity planning has to finish in November for a January fiscal year. Requisition-to-start-date runs 45-75 days for experienced AEs and 60-90 days for entry-level or SDR roles, and that assumes an approved req, an engaged recruiter, and a competitive comp package. Add 30-45 days of onboarding and enablement before the rep touches a real deal. Add 6-9 months to full productivity. The round trip from "we decided we need this person" to "this person is producing at full quota" is nine to twelve months. A Q4 hire is an H2 contributor of the *following* year. That single fact should govern the entire planning calendar.

Cost ranges vary by market and stage, but the structure is consistent. A fully loaded AE costs roughly 1.6-1.9x their base salary once you include benefits, payroll taxes, equity, tooling seats, travel, and the management overhead of one manager per 6-8 reps. On-target earnings for AEs typically run a 50/50 or 60/40 base-to-variable split, and total comp cost as a percentage of the bookings they produce — the comp-cost-of-sales ratio — generally lands somewhere in the 15-30% band depending on segment and motion. Enterprise motions with long cycles sit at the higher end during ramp and drop as tenure builds. If your model produces a comp-cost-of-sales far outside that band, one of your inputs is wrong.

Quota and attainment ranges differ sharply by segment, and this is exactly why blended models fail. Enterprise AEs commonly carry $1.5M-$2.5M with 50-60% attainment and 9-18 month sales cycles. Mid-market lands around $900K-$1.5M at 60-70%. SMB and velocity motions run $600K-$900K at 70-80% attainment with 30-60 day cycles. Notice the inverse relationship: bigger quotas correlate with lower attainment rates, which means a naive blended assumption over-credits enterprise and under-credits SMB simultaneously. Build the model bottoms-up by segment and aggregate, never the reverse.

On tooling cost: below roughly $30M ARR, a disciplined spreadsheet backed by clean CRM exports is genuinely sufficient. The binding constraint at that stage is not modeling sophistication — it is the willingness to type 65% into the attainment cell when the quota letters say 90%. In the $30M-$150M range, purpose-built financial planning tools with pre-built capacity templates start earning their keep because you are coordinating across segments and regions and the spreadsheet version breaks under version-control chaos. Past roughly $100M ARR, enterprise planning platforms with multi-dimensional modeling become the standard, and they carry meaningful annual license costs plus dedicated admin headcount. The failure mode worth naming: buying an expensive planning platform to fix what is actually a discipline problem. The tool will faithfully compute a wrong answer at greater speed and higher cost.

One more timing range that matters: pipeline coverage. Capacity and pipeline are the same problem viewed from two sides. If your enterprise segment historically converts 20-25% of qualified pipeline, you need 4-5x coverage against quota, and that pipeline has to exist *before* the fiscal quarter it closes in — meaning your Q1 capacity plan depends on demand-gen and SDR output from the prior Q3 and Q4. A capacity plan that adds AEs without a corresponding lift in pipeline generation just distributes the same pipeline across more people and drags attainment down for everyone, which is a genuinely destructive outcome: you spend money to make your own attainment metric worse.

Where teams get it wrong

Planning on the average instead of the distribution. The average hides top-rep concentration. When 20% of reps deliver more than half of bookings, the mean is not a description of a typical rep — it is an artifact of a few outliers. Model with quartiles and stress-test what happens when a hire class skews toward the bottom half, which it will roughly half the time.

Under-hiring for attrition. If you plan for 45 reps and hire exactly 45, you are running 32-34 productive reps by Q3 and no recruiting team can close that gap before year-end. The over-hire buffer feels wasteful in January and looks like foresight in August. Model it as insurance with a known premium, not as slack.

Crediting new hires with full quota from day one. Load fifteen hires into Q1, credit them all with full annual quota, and you have overcounted fiscal-year capacity by roughly 30-40% of their aggregate quota. Use the monthly ramp table. Zero out the first 90 days explicitly — not because reps do nothing in their first quarter, but because what they do lands in the *next* quarter's bookings.

Ignoring territory overlap and account coverage. When multiple reps chase the same accounts, effective capacity falls well below nominal capacity through internal competition and duplicated effort. Before you add heads, check whether your CRM shows overlapping account assignment between enterprise and mid-market teams. Adding reps to an over-covered territory does not add capacity; it adds cost and friction. Conversely, if your average book is 300 accounts and reps can meaningfully work 80, you have a coverage problem that more heads genuinely do solve.

Assuming a 40-hour selling week. Quota-carrying reps spend a large share of the week on CRM hygiene, internal meetings, forecast calls, enablement, and administrative work. If the model implicitly assumes forty hours of selling and reality delivers roughly half that, the productivity assumption is inflated at the source. This is also the strongest argument for RevOps investment in automation and admin reduction: every hour returned to selling is capacity you did not have to hire.

Modeling attrition as an even monthly drip. People do not leave uniformly. Departures cluster after commission payouts, after annual quota resets, and after performance reviews — typically producing heavier Q1 and Q3 turnover. Flat 15%-per-year modeling leaves you understaffed exactly when you needed the coverage. Build the attrition curve by quarter and front-load backfill hiring ahead of the predicted spikes.

Treating the plan as a November document. The plan is a hypothesis with a testable prediction. If Q1 actual attainment comes in at 55% against a modeled 65%, you have a ninety-day window to adjust Q3 hiring — and if you wait until Q3 to notice, the window is gone. Instrument the model: track actual attainment, actual ramp times, and actual attrition against the assumptions monthly, and define in advance the threshold that triggers a re-plan.

Letting the CFO negotiate the headcount instead of the assumptions. When someone pushes back on the number, the productive move is to hand them the input sheet. Fewer heads is a completely legitimate choice — as long as everyone signs the sentence that comes with it: "at 32 reps instead of 45, we are modeling $8M of the $30M target as at-risk." Documented trade-offs beat relitigated arguments in Q3.

Choosing an approach: a decision framework

Not every org needs the same depth of model, and over-engineering costs real time. Use company stage and org complexity to pick the approach.

Under ~$20M ARR, single segment, one geography. A single-tab spreadsheet with the core formula, a monthly ramp table, and one attrition assumption is correct and sufficient. Refresh quarterly. Spending three weeks building a scenario-modeling apparatus at this stage is time stolen from fixing the CRM data that the model depends on. Priority order: clean attainment data first, then the model.

$20M-$75M ARR, two or three segments. Move to a segmented bottoms-up model — one block per segment, each with its own quota, attainment, ramp, and attrition, aggregating to a total. Add two stress scenarios. This is where most RevOps teams should live, and where the biggest single improvement is usually replacing a blended attainment number with three segment-specific ones.

$75M+ ARR, multiple regions, multiple products. Now you need real scenario modeling and version control, because territory-level plans have to roll up cleanly and finance needs to see the capacity plan reconcile against the operating plan without manual re-keying. This is the point where dedicated planning software earns its cost, and where a named owner for the model becomes a real part of someone's job rather than a November side quest.

The second decision — how aggressive to be on the over-hire buffer — turns on how much you trust the demand side. If pipeline coverage is healthy and demand-gen is hitting plan, hire toward the top of the buffer range; the constraint is selling capacity and more reps convert to bookings. If coverage is thin, hiring more reps makes attainment worse and morale worse simultaneously, and the money is better spent on demand generation. Run this check before you finalize: divide projected pipeline by projected capacity-quota. If the ratio is below your historical conversion requirement, you have a demand problem wearing a capacity costume.

The third decision is build-versus-borrow for coverage gaps. When the model shows a shortfall that recruiting cannot close in time, the options are a partner or channel motion, contractor or fractional sellers for a defined window, redistributing accounts toward proven reps, or simply accepting a lower target for that segment. All four are defensible. What is not defensible is leaving the gap in the model unlabeled and hoping the year is kind.

Related questions

How many reps can one sales manager support?

Common spans run six to eight quota-carrying reps per front-line manager, tightening toward five or six in complex enterprise motions and stretching to ten in high-velocity SMB teams. Your capacity plan must include the management headcount those spans imply, or you fund reps with nobody to coach them.

Should capacity planning be owned by RevOps or Finance?

RevOps owns the model and the operational inputs — attainment, ramp, attrition, territory data. Finance owns the target, the cost envelope, and the approval. Joint ownership with a single named modeler works best; two competing spreadsheets is the failure state.

How do you plan capacity when the fiscal year target isn't final?

Model three demand scenarios against the same productivity assumptions and identify the hiring decisions that are common to all three. Those are safe to start recruiting for immediately. Hold the divergent hires until the target locks — recruiting lead times mean the common set is often most of the plan.

Does capacity planning change for a usage-based or consumption revenue model?

The structure holds but the unit changes: you are planning capacity against net new consumption commitments and expansion rather than fixed-ARR bookings. Attainment tends to be more volatile, so widen your scenario range and lean harder on the over-hire buffer.

FAQ

What is the single most common mistake in sales capacity planning?

Assuming reps will hit quota. Real attainment across a full sales org typically lands in the 50-75% range, and planning on heroic productivity produces systematic understaffing. The correction is mechanical: use trailing-twelve-month actuals from your own CRM, at the rep level, including the underperformers.

How should I handle ramp time for new hires?

Build a monthly ramp table from your own last three hire classes rather than applying one blended factor. The typical shape is near-zero contribution in the first 90 days and full productivity at six to nine months, netting roughly 60-75% of quota over the first twelve. Start month matters enormously, so model each hire's start date individually.

What minimum data do I need to start?

Segmented bookings target, quota per rep by segment, trailing-twelve-month attainment distribution, cohort ramp curve, and attrition rate from HRIS. If you are missing the attainment history, that is your first project — every other input is guessable and that one is not.

How often should the capacity plan be revisited?

Track drift monthly, formally review quarterly, and define a specific trigger in advance — for example, Q1 attainment landing more than ten points below model automatically opens a re-plan for H2 hiring. Annual-only planning means discovering a gap when it is too late to hire against it.

What do I do when the CFO wants fewer heads than the model says?

Hand over the input sheet and make them choose which assumption changes. If the headcount is genuinely capped, quantify the shortfall explicitly and put it in writing: at reduced capacity, state the specific dollar amount of the target that is now at risk. That converts a judgment call into a documented, jointly-owned trade-off.

Can I use the same formula across SDRs, AEs, and CSMs?

Yes — demand divided by throughput times realization times ramp is universal. Only the units change: qualified opportunities for SDRs, bookings for AEs, accounts or renewal dollars for CSMs. Run each role as its own model with its own inputs, then check that the roles are proportionally sized to each other.

Sources

• https://www.bridgegroupinc.com/blog
• https://www.saastr.com/
• https://openviewpartners.com/blog/
• https://www.alexandergroup.com/insights/
• https://www.mckinsey.com/capabilities/growth-marketing-and-sales/our-insights
• https://hbr.org/topic/subject/sales
• https://www.salesforce.com/resources/research-reports/state-of-sales/
• https://www.gartner.com/en/sales
• https://www.anaplan.com/blog/

Related on PULSE

• How Do I Build a Sales Capacity Plan to Hit Next Year's Number in 2027?
• What is capacity buffer math for sales planning in 2027?
• How do you model capacity planning for SDR hiring against pipeline targets?
• How do 2027 longer sales cycles impact your quota capacity model for enterprise AEs?
• How do you build a capacity model for sales hiring in 2027?
• How should a 2027 sales org plan value-engineering capacity for enterprise deals?