Most finance teams treat sales compensation accruals as a routine close task. Pull bookings, apply a blended commission rate, layer in an attainment factor, post the journal entry, move on.
That process works on paper. In practice, the resulting accrual is almost always wrong, and it is almost always wrong in the same direction. Most companies under-accrue commissions by 5 to 15 percent in any given month, which means the comp expense reported during close does not match the comp expense that eventually gets paid out.
The variance is structural. It is not caused by data quality issues or analyst error. It is caused by gaps in the model itself — gaps that no one notices because the standard accrual formula was inherited from someone who left three years ago and nobody has touched it since.
This article unpacks why the standard accrual model is wrong, what the structural causes of variance are, and what a tighter framework looks like. It is written for a finance leader who has noticed that the comp accrual line keeps drifting and wants to do something about it without rebuilding the entire compensation plan. It sits alongside two related CFO-focused questions about sales compensation — sales compensation as gross margin engineering and sales rep ramp economics — that together cover the design, forecasting, and modeling discipline most companies are missing.
Why accrual accuracy matters more than people think
A few percentage points of monthly variance on the comp line sounds tolerable. It is not.
Sales compensation is usually the largest variable expense in a SaaS or B2B company below cost of revenue. A 10 percent miss on a $2 million monthly comp accrual is a $200,000 variance, repeated twelve times a year. That kind of consistent drift shows up in three places that matter:
Monthly close quality. Persistent accrual misses force true-up entries that distort the comparability of monthly results. The CFO has to explain why operating income jumps in some months and dips in others, when the underlying business performance was flat.
Board-level forecasting credibility. If the monthly comp expense is consistently 5 to 15 percent below what gets paid out, the rolling forecast inherits the same error. The CFO presents a plan that misses on the comp line every quarter, and the board notices.
Audit comfort. Material variance between accrued and paid comp invites audit scrutiny, especially under ASC 606 where the relationship between commission expense and revenue recognition matters. A clean accrual is not just an operational nicety. It is part of the audit defense.
The good news is that the structural causes of variance are well understood. The bad news is that the standard accrual model is not designed to capture any of them.
The standard accrual model and what it misses
The model in most companies looks something like this:
Accrued commission expense = Monthly bookings × Blended commission rate × Historical attainment factor
Sometimes the formula is slightly more elaborate, splitting by team or product. The core structure is the same: a blended rate applied to a bookings number, adjusted by a single attainment multiplier.
This model is convenient because it can be calculated quickly from a CRM bookings report and a comp plan summary. It is wrong because it embeds at least five assumptions that almost never hold:
- The blended rate is stable across deals
- Attainment is linear and predictable
- The plan does not change mid-period
- Late adjustments and clawbacks are immaterial
- Bookings and payout timing are aligned
Each of these assumptions creates a structural source of accrual variance. Let us walk through them in order.
Cause 1: Mid-period plan changes that the accrual never sees
Most sales compensation plans are not static. SPIFFs get introduced mid-quarter. Accelerator thresholds get adjusted after a slow start to the year. New product incentives roll out alongside product launches. Team-based bonuses get layered onto existing individual plans.
Each of these changes is communicated to the sales team. None of them get communicated to the accrual model. The blended rate sitting in the accrual formula was calibrated at the start of the year and has not been touched since.
This is the single largest source of accrual variance in most companies. The fix is procedural rather than mathematical: any change to the comp plan that affects expected payout — including overlays, SPIFFs, and adjustments — needs to flow through to whoever owns the accrual formula. In most organizations this is a missing handoff between RevOps and Accounting that nobody has ever explicitly assigned.
Cause 2: Accelerator non-linearity
Most comp plans pay accelerated rates above quota attainment thresholds. A rep at 95% of quota and a rep at 105% of quota have very different marginal payouts on their next deal, even though their attainment differs by ten points.
The standard accrual applies a single blended attainment factor across the entire team. This works when the team’s attainment distribution is symmetric around the mean. It fails when the distribution is skewed, which it usually is in any quarter with strong or weak overall performance.
In a strong quarter, more reps cross into accelerator territory than the model assumes. The accrual under-states the actual payout. In a weak quarter, fewer reps reach accelerators, but the model has already baked in an attainment factor close to the historical average. The accrual over-states.
The fix is a rep-level or cohort-level attainment model rather than a team-blended one. The math is not complicated. Run the accrual at the cohort level (top quartile, middle two quartiles, bottom quartile) and apply the appropriate accelerator multiplier to each. Most of the non-linearity error disappears.
Cause 3: The ramp tax
New hires earn comp at rates that are systematically misaligned with their productivity during their ramp period. They often have a ramped quota that scales up over two to three quarters, but they earn full base salary, plus draws or guarantees in the early months that pay out regardless of performance.
The accrual model usually treats new hires the same as ramped reps. It applies the team’s blended attainment factor to the new hire’s small bookings number, producing a tiny accrued commission amount. Reality is the opposite: the new hire is earning closer to full target comp through guarantees while producing minimal bookings.
The variance per new hire is not huge, but most growth-stage companies are hiring continuously. Across a sales org with twenty new hires in their first two quarters, the cumulative under-accrual can run into six figures per month.
The fix is to model ramping reps separately from fully ramped reps. Their accrued commission expense should be closer to their target OTE pro-rated to the month, with the bookings-driven calculation only kicking in once they exit ramp.
Cause 4: Late adjustments and clawbacks
Comp does not settle at booking. It settles weeks or months later, after the deal is processed, the revenue is recognized, the disputes are resolved, and any clawback windows have closed. The underlying commission calculation error rate is also part of this drag — every disputed payout that gets corrected weeks later is variance the accrual did not anticipate.
The standard accrual treats the booking event as the trigger and applies the full expected commission. In reality, a meaningful portion of bookings result in adjustments — partial cancellations, deal restructures, customer disputes, revenue rec delays, or outright clawbacks. The actual payout is typically 92 to 97 percent of the booking-implied amount.
The fix here is a tracked adjustment factor based on historical experience. A monthly accrual that applies a 95 percent realization factor to booking-implied commission will under-accrue less often than one that assumes 100 percent realization. The exact number depends on the business and should be re-calibrated each year.
Cause 5: Bookings and payout timing misalignment
Most comp plans do not pay on bookings alone. They pay on a mix of triggers — bookings, billings, invoicing, customer payment, revenue recognition, or specific milestone events. The accrual model usually assumes commission expense matches bookings timing because the bookings number is what the analyst has.
When the payout trigger lags bookings, the accrual lags the actual payout liability. For ASC 606 purposes the timing matters because commission expense is supposed to be matched to the revenue stream it generated. Mismatched accrual timing creates audit-visible inconsistencies.
The fix is to align the accrual timing with the actual commission earn date, not the booking date. This is more work to set up, but once configured the accrual becomes structurally cleaner.
A tighter accrual framework
A more accurate accrual model has four components:
Cohort-based attainment. Replace the single team-blended attainment factor with cohort-specific factors for top, middle, and bottom performers. This eliminates most of the accelerator non-linearity error.
Ramped vs. fully ramped reps. Model new hires on a target-comp basis during their ramp period and switch to bookings-driven calculation only after they fully ramp. This eliminates most of the ramp tax error.
Tracked realization factor. Apply a historically calibrated realization factor (typically 92–97%) to account for adjustments, partial clawbacks, revenue rec delays, and other post-booking commission reductions.
Plan change tracking. Establish a quarterly review where any mid-period plan changes — SPIFFs, accelerator adjustments, overlay programs — get reflected in the accrual model parameters. This is procedural, not mathematical, but it is the single largest source of variance.
A model that does these four things typically reduces monthly accrual variance from 5–15% to under 2%, which is in the noise band of any reasonable forecasting process.
What to do about residual variance
No accrual model will be exact. The goal is not zero variance. It is variance that is structurally explainable and small enough that it does not distort the monthly close.
The remaining residual variance should be tracked and analyzed. A consistent pattern in residuals — for example, the accrual always under-shoots in Q4 — usually points to a model gap that has not been closed. A noisy residual with no clear pattern is acceptable.
Most CFOs find that simply having a model that explains its own variance is enormously valuable. When the comp accrual misses by 4% in March and the controller can point to specific causes — three SPIFFs introduced mid-quarter, one accelerator threshold change, two cohort attainment surprises — the conversation with the audit committee becomes very different than when the answer is “we are not sure.”
Final takeaway
Sales compensation accruals are not a routine close task. They are a structural finance modeling problem that most companies have not solved because the standard model was inherited rather than designed.
A tighter accrual framework will not change the total amount of commission the company pays. It will change the monthly distribution of that expense, the smoothness of the income statement, and the credibility of the rolling forecast. For a CFO who closes the books every month and presents a forecast every quarter, that is not a marginal improvement. It is a meaningful upgrade to two of the most visible parts of the finance function.
The fix is not more complicated math. It is a model that understands the actual mechanics of how sales compensation gets paid, and a procedural connection between RevOps and Accounting that makes sure mid-period changes do not get lost.
Frequently Asked Questions
How much accrual variance is typical?
In our experience working with mid-market and enterprise companies, monthly sales comp accrual variance commonly runs in the 5 to 15 percent range. Companies with tighter integration between RevOps and Accounting, and explicit modeling of ramp and attainment cohorts, can reduce this to under 2 percent.
Is this a problem at small companies too?
It is a smaller dollar problem but a larger percentage problem. Small companies have noisier comp expense lines because individual rep outcomes drive a larger share of the total. The same structural causes apply, but the variance is often masked by overall noise in the financials.
Does this affect ASC 606 deferred commission accounting?
Yes. ASC 606 requires commission expense to be matched to the related revenue stream over the customer life. The accrual model feeds into the capitalization and amortization calculation. A consistently inaccurate accrual creates downstream inaccuracies in the capitalized commission asset and the amortization schedule.
Who should own the accrual model?
Accounting owns the final journal entry, but the model parameters should be owned jointly with RevOps. The most common gap is that Accounting owns a model whose inputs depend on plan information that only RevOps has visibility into. Bridging that gap is more important than picking which department owns the model.
How often should the model be recalibrated?
The realization factor and cohort attainment parameters should be recalibrated annually. Mid-period plan changes should flow through to the model whenever they happen, which in practice means at least quarterly. The cohort definitions themselves rarely need to change unless the sales org structure changes materially.
What is the single highest-impact fix?
For most companies, the single highest-impact change is connecting mid-period plan changes to the accrual model. This is a procedural fix, not a modeling change, and it usually accounts for more than half of the structural variance in the typical company.
