ASC 606 went effective for public companies in 2018 and for most private companies by 2020. For accounting, the transition meant rethinking when revenue gets recognized — not when cash arrives or when an invoice is sent, but when performance obligations are satisfied. For FP&A, the transition created a problem that many teams are still quietly managing around: the budget was built on intuition about when money comes in, while the actuals are reported based on when obligations are fulfilled. In contract-heavy or subscription businesses, those two timing assumptions diverge enough to generate systematic variance noise that analysts spend meaningful hours chasing every close cycle.
What ASC 606 Changed for Revenue Timing
Before ASC 606, many companies recognized revenue on a completed-contract or milestone basis — often closely aligned with when they expected to receive cash. The new standard requires recognizing revenue as performance obligations are satisfied, which can be either at a point in time or over time, depending on the nature of the contract. For a SaaS company, that's typically straight-line over the contract term. For a professional services firm with milestone deliverables, it might follow a percentage-of-completion method. For a hardware-plus-support bundle, it requires an allocation of the total contract price across distinct performance obligations using stand-alone selling prices.
Each of these patterns creates a timing relationship between cash receipts and revenue recognition that is specific to the contract structure. And FP&A teams that built their budgets using cash-receipt timing as a proxy for revenue timing — which is the intuitive approach when you're modeling a business in a spreadsheet — end up with a structural mismatch. The actuals follow the obligation curve; the budget follows the billing calendar.
That mismatch doesn't show up as a single large variance. It shows up as persistent, small-to-medium variances in the revenue line that move in direction and magnitude based on the composition of the current period's contract mix. A period with a high proportion of new long-term contracts may show below-budget revenue recognition even when bookings are above budget. A period with a high proportion of contract completions may show above-budget recognition even when new sales were soft.
The Budget-to-Actuals Comparability Problem
The deeper issue is comparability. A variance is only analytically meaningful if the budget and the actuals are measuring the same thing. For revenue under ASC 606, budget and actuals are often measuring different things — not because of an accounting error, but because the budget was constructed at a level of abstraction that didn't attempt to model recognition timing at the contract level.
Consider a professional services firm that budgeted Q3 revenue at $2.4M based on expected project completions and billing milestones. Actuals came in at $2.1M. The revenue variance of $300K unfavorable is real — but it breaks into three components: $180K is a genuine delivery delay on a project that slipped to Q4 (true operational miss); $75K is a recognition timing difference where a milestone was completed but the billing hadn't yet been sent so the revenue recognition was deferred to the following period under the company's ASC 606 policy; and $45K is a contract modification that changed the allocation of contract price across obligations, shifting some recognition from Q3 into the full contract term.
The FP&A analyst who reports "$300K revenue miss" to the CFO is technically correct. The CFO who asks "is this a sales problem or a delivery problem?" will get an incomplete answer if the analyst doesn't have the decomposition. The board that hears "$300K unfavorable, primarily delivery timing" gets a much less useful picture than the board that hears "$180K delivery delay + $75K billing-timing recognition deferral + $45K contract modification reallocation." The first framing implies a Q4 catch-up. The second explains which of the three components are likely to reverse, and on what timeline.
How the Mismatch Compounds Over Time
One-off timing differences are manageable if they're identified and tracked. The problem is that most FP&A processes don't systematically capture recognition timing differences as a distinct variance category — they treat them as part of the overall revenue variance, which gets explained in aggregate terms. Over several periods, this creates a noise floor in the revenue variance analysis that the team learns to expect but never fully explains. The variance "around" the timing issue becomes normalized.
For SaaS businesses with ARR-based metrics, the timing misalignment creates a different problem. If the budget was built with straight-line ARR recognition and the actuals reflect contract-specific recognition curves (annual contracts that start mid-month, multi-year deals with ramps, usage-based components that vary monthly), then ARR actuals will frequently diverge from the budget in ways that don't reflect the health of the subscription base. A cohort of new contracts signed in the last week of the period will under-recognize relative to budget for that period but then catch up in the following month. An analyst reviewing monthly ARR actuals against a straight-line budget will see unexplained volatility that is entirely an artifact of recognition timing, not a change in the business.
We are not suggesting that FP&A teams should rebuild their entire budget on ASC 606 recognition curves — that would add significant complexity to the budget process for marginal analytical gain. What we are suggesting is that the variance analysis layer should include a recognition-timing reconciliation that identifies, at each close, how much of the revenue variance is attributable to recognition timing rather than genuine business performance. That reconciliation converts unclear noise into a classified, trackable variance type — which is categorically more useful for the board narrative and for the next period's forecast.
Practical Steps Toward Cleaner Revenue Variance
The intervention points are in the budget construction process and in the close analysis process:
Budget side: For any contract type that represents more than 15% of total revenue, the budget should model recognition timing explicitly — not necessarily at the contract level, but at the cohort level. New contracts starting in month one of the quarter should be budgeted with the partial-period recognition haircut that the accounting policy will apply. Annual renewals should be timed to match the billing anniversary, not the calendar quarter.
Close analysis side: The revenue variance should be structured as: (1) volume variance — contracts won/lost vs. plan, (2) rate variance — average contract value vs. plan, (3) recognition timing variance — the difference between when cash was expected and when the obligation satisfaction curve recognized it. The third component is a reconciling item, not a business performance metric. Separating it from the first two prevents the CFO and board from drawing operational conclusions from accounting-timing noise.
Revenue analysis gets sharper when the analyst can say: "The volume and rate variance together explain $120K of the miss. The remaining $75K is recognition timing and is expected to appear in Q4 actuals as those obligations complete." That is the kind of precision that builds CFO and board confidence in the FP&A team's understanding of the business — not because it makes the number look better, but because it demonstrates that the team knows the difference between a signal and a timing artifact.
