It's February 15. The fiscal year started 6 weeks ago. Your CEO asks when the year's budget will be approved. The truth: your team is on the fourth round of revisions, functional leaders are debating numbers, and the board wants "a more conservative version."
The AOP approved in February is no longer a plan — it's archive. The capex, hiring, and commercial commitments of the current year were already made without it. Operations runs by inertia while finance debates numbers.
The problem is not lack of talent or time. It's process design. Companies that close AOP in November aren't faster — they have a disciplined calendar with specific bottlenecks eliminated. This module teaches that calendar.
By the end of this module you'll have it down. Let's go.
Your CEO tells you: "we need a more realistic AOP. Let's do a fifth round of revisions." What do you answer?
In plain language
Before the mechanics, the four basic questions.
Why do this at all?
So the company starts the fiscal year with a shared contract among all operational leaders. AOP = the "what we commit to do" of the year. Without an approved AOP before starting, commitments are individual and uncoordinated. With approved AOP, everyone fights on the same team against the same number.
Who builds it?
FP&A coordinates. CFO leads the process. Functional leaders (Sales, Operations, Marketing, IT, HR) build their pieces. CEO arbitrates when there are gaps. Board approves the close. Five roles, five responsibilities — without role clarity, AOP gets stuck.
When is it done?
Ideal calendar: kickoff mid-July (board strategic posture), bottom-up from functions August-September, arbitration October, lock November, announcement + communication December. AOP approved and published BEFORE January 1. Anything after is archive, not plan.
What if we close it late?
Operations makes current-year decisions without AOP. Capex gets approved ad-hoc, hires happen by functional pressure, commercial commitments without frame. When AOP finally gets approved, it's already contradicted by decisions taken. Loses its function as shared contract.
Andina S.A. — the AOP that took 5 months
When the new CFO arrived at Andina, he found the typical pattern: AOP approved in March, with the fiscal year already underway. The internal excuse: "it's a complex business, needs iterations." The reality: 5 rounds of revision without cap, no formal arbitration, no board strategic posture before starting.
The finance team spent 5 months on AOP. Then spent the year explaining variances against a plan no one believed when it was approved. Operations, meanwhile, had made decisions in January-February without frame.
In the new CFO's first cycle, he reduced AOP to 10 weeks: board strategic posture in July (BEFORE the calendar), single template bottom-up August-September, 2-week arbitration committee in October, lock first week of November. AOP approved by board November 15. It worked.
The visual below shows exactly this calendar, with the 5 bottlenecks that stretched Andina's AOP. Turn each off and watch the calendar contract from 20 weeks to 8.
AOP calendar live
Top-down and bottom-up run in parallel. The calendar doesn't fail by number of weeks — it fails by broken convergence. If the two tracks don't meet in arbitration, there's no lock.
Turn off each bottleneck and watch: weeks contract, lock approaches, and AOP finishes before January 1 — as it should.
Interactive visual
AOP — the calendar that almost always fails
Top-down and bottom-up run in parallel from July. They converge at the November lock point. Bottlenecks stretch the calendar or break the convergence. Turn each off and watch the plan adjust.
17weeks
Active bottlenecks
What you are seeing
AOPs that do not finish on time fail not from numbers but from process design. The 5 bottlenecks here are universal in mid-market. Companies that close AOP in November have all 5 resolved. Companies that close in February (year already started) have 3+. The discipline isn't complicated — it's choosing to resolve them vs tolerate them. The difference between an AOP that serves management and one that is just exercise: the first is approved and published before January 1.
Mental rule: the AOP calendar is NOT flexible. It's 8-10 weeks or it doesn't work. When a leader asks for "more time," the right question is "what specific bottleneck is stretching this?" — not "how many more weeks?" Resolving the bottleneck is 10x more effective than extending the calendar.
Once AOP is closed, it stops iterating. Deviations are managed via Rolling Forecast (Module 2.3) which updates monthly without touching the approved AOP. AOP is anchor; RAF is compass.
The mechanics: how to design an AOP that closes on time
- Board strategic posture BEFORE starting. 4-6 hour session in July to align defend/extend/disrupt. Without this, AOP reflects individual biases, not direction.
- Single template, monthly granularity. FP&A publishes template by end-July. Revenue, COGS, opex, headcount, capex by month. Any leader who invents their own format generates extra work for FP&A.
- Explicit cap on iteration rounds. Maximum 2: bottom-up draft → arbitration → final. More rounds = process failure. If leaders don't converge in 2, goes to formal arbitration committee.
- Arbitration committee with short mandate. CEO + CFO + COO. 2 weeks to resolve material gaps. Final decision, not consultative. Without this, disagreements go forever.
- Public calendar with consequences. Hard deadlines: late submission = use prior year + 5%. Sounds drastic — works. Without consequence, deadlines are suggestion.
- Post-approval communication. Once the board approves, all leaders present THEIR piece of AOP to their team in December. AOP is not finance's property — it's everyone's contract.
- LRP (Long Range Plan). 3-5 years. Strategic. Qualitative + quantitative. Done once a year. Anchors direction. (Module 2.1)
- AOP (Annual Operating Plan). 1 year. Operational. Monthly granularity. Approved before the fiscal year. The year's execution contract.
- Rolling Forecast. 12-month rolling, updated each month. Reflects updated reality. Does NOT replace AOP — informs when AOP is off-track and how much. (Module 2.3)
- Confusing the three is the most common error. AOP that "updates every month" stops being AOP — loses its anchor function. Rolling Forecast pretending to replace AOP eliminates the year's execution contract.
Adversarial check
Adversarial check
1.Your Sales lead says: "I can't deliver my AOP piece in September, I need until November because the market is uncertain." What do you answer?
2.Mid-year the AOP is 15% below approved. Your team says: "we need to redo AOP to reflect reality." What do you decide?
3.What is the most important difference between an AOP and a Rolling Forecast?
Exit checklist
Suggested re-review: every year, after AOP close, post-mortem with FP&A to identify which bottleneck to eliminate next cycle.
Optional
Go deeper
Sources and books to dig into the original material