Month-end accruals for service spend: a CFO guide

Executive Summary

Service spend accruals fail differently than product cost accruals. A purchase order tells you what a shipment should cost. A staffing contract, a freight lane, or an IT services statement of work tells you what an invoice should cost only if someone checks the invoice against the contract terms, and most close calendars do not build in time for that check.

The mechanism is simple: AP estimates the accrual from last month's invoice or a run-rate average, not from the contract. When the actual invoice arrives with a surcharge, a rate escalation, or a missed rebate, the variance shows up as a true-up in a later period, not as an explained item in the period it belongs to.

What changes this is moving the accrual basis from "what we paid last time" to "what the contract says we owe," which is a contract compliance exercise finance teams rarely staff for. The fix is not a bigger accrual reserve. It is a faster, contract-based estimate.

1. Why do service spend accruals miss more often than product cost accruals?

Product costs are governed by a purchase order with a fixed price and quantity, so the accrual is close to mechanical. Service spend is governed by a contract with variable terms: rate tiers, surcharges, minimum commitments, and rebate thresholds that depend on volume nobody has tallied yet at close. AP accrues from the last invoice because reading the contract takes time the close calendar does not allow, so the estimate inherits whatever error was already in that last invoice.

A freight invoice carries a base rate plus accessorials that change with fuel prices and lane volume. A staffing invoice carries a bill rate plus overtime and shift differentials that depend on hours nobody has reconciled by close. Both require someone to hold the contract next to the estimate, not just the prior invoice.

When the estimate is copied from last month, any error in last month repeats itself: a surcharge that started incorrectly stays in the accrual until someone opens the contract and checks it against the invoice.

A standard close checklist has no step that asks whether the contract terms still match what is being estimated. Without that step, the accrual carries the same assumption forward until an audit or a variance review forces a look at the underlying contract.

2. What does this do to the gross margin line the board sees?

An accrual miss on service spend shows up as a margin swing in the period the true-up lands, not the period the drift happened. A board sees a one-month dip or lift with no clean explanation, because the finance narrative for that month has to describe a prior period's error rather than that period's operations. This is the exact mechanism behind an unexplained gross margin gap: a real cost event, recognized late, in the wrong period.

A CFO presenting a margin bridge wants every line item traceable to a cause: volume, mix, price, or cost. A true-up from a stale accrual does not fit cleanly into any of those buckets, so it gets labeled as a one-time item or folded into cost of goods without explanation.

Over several quarters this produces a pattern the board notices even without knowing the mechanism: margin that moves in ways operations cannot account for. That pattern is harder to explain than a single bad quarter, because it suggests the finance function does not have a handle on service spend.

The repair is not a better bridge chart. It is closing the gap between what gets accrued and what the contract actually specifies, so the true-up stops recurring.

3. Which service categories create accrual risk at close?

Freight, contract labor, and IT or professional services each carry accrual risk because each has a contract structure with a variable component that does not appear on a simple invoice total. Each requires its own accrual method rather than one blanket run-rate approach, and the categories below are described on their own mechanics rather than ranked against each other for size or frequency.

Each category fails at close for a different structural reason, which is why one accrual method does not serve all three.

A. Freight and 3PL

Base rates are stable within a contract period, but fuel surcharges and accessorials move with published indices and lane-level activity. An accrual built on last month's total invoice carries forward any surcharge that was applied against the wrong reference table, and the error does not surface until someone checks the surcharge condition itself.

B. Contract labor and staffing

Bill rates are contractual, but hours, overtime, and shift differentials are operational and not always reconciled by close. A not-to-exceed cap adds a second failure point: the accrual can be correct on rate and still miss a cap breach that the invoice, once it arrives, does not respect.

C. IT and professional services

Statements of work often bill on milestones or blended rates that do not map cleanly to a calendar month. The accrual has to estimate percent complete on work finance cannot directly observe, which is a different problem from a surcharge or a bill rate and needs its own estimation method.

4. How should a CFO change the accrual method without slowing the close?

The fix is not adding review time to every invoice at close. It is building the accrual formula once, per contract, so the monthly estimate references the rate card and surcharge conditions directly instead of the prior invoice. Once that formula exists, applying it at close takes the same time as the run-rate method it replaces, because the judgment work happened once, up front, not every month.

Building the formula means someone reads the contract once: the rate tiers, the surcharge triggers, the rebate thresholds, and the caps, and turns each into a rule the close process can apply. That is a one-time cost per vendor contract, not a recurring one.

After that, the monthly accrual is a lookup against the rule set rather than a copy of last month's number. The close calendar does not lengthen, because the estimate is now a calculation instead of a judgment call made under time pressure.

This is the same work a contract compliance audit does for historical invoices, applied prospectively to the accrual instead of retrospectively to a paid invoice. The two are the same skill set aimed at two different time horizons.

5. Is this worth fixing given everything else on a CFO's plate at close?

A stale accrual method is a recurring exposure, not a one-time rounding error, because the same contract terms get misread every close cycle until someone corrects the formula. The question is not whether it is worth fixing but whether it gets fixed before or after it shows up as an unexplained board question about margin. Once the rule set exists, the exposure closes for every future close, not just the current one.

A CFO weighing this against other close priorities should separate two costs: the accounting cost of a wrong accrual, which is a true-up and a restated variance, and the narrative cost, which is a board conversation about why margin moved without an operational cause.

The accounting cost is bounded and correctable. The narrative cost compounds, because a board that has seen one unexplained margin swing scrutinizes the next one harder, whether or not it is related.

Fixing the accrual method is a fixed piece of work per contract, not an ongoing burden. It is the same category of investment as a first-100-days close diagnostic for a new finance leader: done once, it changes the baseline for every close after it.

6. What does a CFO actually ask AP to change starting next close?

Ask AP to accrue from the contract's rate schedule and surcharge conditions for the highest-spend service vendors first, not all vendors at once. Start with freight, contract labor, and any vendor with a not-to-exceed cap, because those three carry variable terms per invoice that a run-rate estimate cannot capture. Everything else can stay on the existing run-rate method until the highest-risk vendors are converted.

This is a sequencing decision, not an all-or-nothing policy change. A CFO does not need every vendor converted to contract-based accrual by next close; a small set of high-spend, high-variability vendors will carry the bulk of the exposure.

AP needs the rate schedule and surcharge terms in a usable form, which usually means someone extracting them from a PDF into a rule the accrual template can reference. That extraction work is the same work a contract compliance review performs, just aimed at the accrual instead of the paid invoice.

Once the top vendors are converted, the true-up variance on those lines should shrink within one or two close cycles, giving the CFO a concrete before-and-after to point to when explaining the change.

For the wider pattern this sits inside, start with the margin drift guide.

For the wider pattern this sits inside, start with the margin drift guide. See also the six categories drift hides in and margin drift vs. legitimate price increases: how to tell them apart.

7. Frequently Asked Questions (People Also Ask)

What is a service spend accrual?

It is an estimate, booked at month end, of what a company owes a service vendor for work or delivery that already happened but has not yet been invoiced. Unlike a product cost accrual tied to a purchase order price, a service accrual has to estimate against contract terms like rate tiers and surcharges that may not be fully known until the invoice arrives.

Why does a service accrual true-up show up months later?

Because AP typically books the accrual from a prior invoice or a run-rate average rather than from the contract itself. When the real invoice arrives with a different rate, surcharge, or rebate application, the difference between the estimate and the actual gets corrected in the period the invoice lands, not the period the cost was incurred.

Does a not-to-exceed cap affect how an accrual should be built?

Yes. A not-to-exceed cap limits what a vendor can bill regardless of hours or usage. An accrual that only estimates the bill rate and hours can still miss a cap breach, because the cap is a separate contract condition that has to be checked against the estimate independently.

Should the accrual formula be rebuilt every month?

No. The contract terms, rate tiers, surcharge triggers, and caps only need to be extracted into a rule once per contract. After that, each month's accrual is a lookup against the existing rule set rather than a fresh read of the contract, which keeps the close calendar the same length.

Who should own fixing the accrual method, controller or AP?

The controller typically owns the accrual policy and its accuracy, while AP executes the monthly calculation. Building the contract-based rule set is closer to a controller or contract compliance function's work, since it requires reading and interpreting contract language, not just processing invoices.

How is this different from a general AP recovery audit?

A recovery audit looks backward at invoices already paid to find overbilling or missed credits. Fixing the accrual method is a forward-looking application of the same skill: reading the contract once and turning its terms into a rule, but applied to the estimate before the invoice arrives instead of to the invoice after it is paid.

What happens if a company just increases its accrual reserve instead of fixing the method?

A bigger reserve can absorb the size of the miss but does not fix the timing or the explanation problem. The true-up still lands in the wrong period and still shows up as an unexplained margin swing; a larger cushion just makes it less likely to be visible to the board, not less likely to happen.

Which vendor contracts should a CFO prioritize converting first?

Start with the highest-spend service vendors that have variable contract terms: freight and 3PL contracts with surcharges, staffing contracts with overtime or not-to-exceed caps, and any statement of work billed on milestones. Lower-spend, fixed-fee vendors carry less accrual risk and can stay on the existing method longer.

Can this be done without adding headcount to the close process?

Yes, because the work of reading the contract and building the rule happens once, outside the close calendar, rather than every month during close. Once the rule exists, applying it at close is a lookup, which takes about the same time as the run-rate method it replaces.