The 2-Hour Committee That Cost R$1.8M/Year in Delayed Decisions
12 leaders, 50 weeks, R$360k in senior payroll in the room alone. The pyramidal cascade below summed another R$4.3M/year. A War Story from a B2B SaaS.
Before we start: this is a model, not a client dossier. The numbers come from public research (Microsoft, Atlassian, Harvard, Rogelberg) applied to the fully loaded hourly cost of a mid-sized Brazilian SaaS. Treat every line as a calculation you redo with your own figures, not as a promise from us.
The scenario
A Brazilian B2B SaaS, fintech, 600 people: 12 in leadership, 60 in middle management, 528 between senior specialists and analysts. Post-Series B, ARR around USD 38M. A veteran CFO, a COO just arrived from a US company. The board pressing for a Rule of 40 above 40% by year-end, and the math would not close: revenue grew 28% over the last year, payroll grew 31%. When headcount rises faster than ARR, operating leverage walks backward, and everyone feels it, but nobody points to the spot. The CFO knew there was a hole. He did not know where, or on which floor.
The symptom nobody called a symptom
Every Thursday, from 2 to 4 PM, the "Operational Executive Committee" ran. Twelve chairs filled: CEO, COO, CFO, CTO, CPO, the VP of sales, the heads of CS, marketing, people, and legal, plus two product leads. The usual agenda: OKR review, status of cross-area initiatives, pending decisions, risk read. It looked reasonable. Every inherited meeting looks reasonable, until someone puts a clock beside it and adds up who is sitting there.
The math on the hour is plain: 12 leaders at R$ 240k/year of loaded cost, divided by 1,920 working hours, times 2 hours, times 50 weeks. That comes to R$ 360k/year just to be present in that room. That is the easy number, the one that fits on the leadership payroll line and that any spreadsheet can see.
The hard number showed up when the COO went after what the Thursday committee pushed downward. Because the committee did not happen on its own. It pulled an entire pyramid of people busy preparing to feed it.
The layer nobody measured: the pyramid under the committee
Each of the 12 leaders on the Thursday committee had their own weekly ritual to "prep the material". The pattern repeated:
- Tuesday, the area prep meeting: each leader gathered the team for 90 minutes, 6 people on average, to assemble what would go up.
- Wednesday, the loose alignments: 2 to 3 one-on-one or small-group conversations, 30 minutes each, to close loose ends.
- Friday, the cascade meeting: another 45 minutes with the team, 8 people on average, to relay what the committee decided.
| Layer | Activity | Person-h/year | Cost R$/h | Annual cost R$ |
|---|---|---|---|---|
| Leadership | Thursday committee, 2-4 PM | 1,200 | R$ 300 | R$ 360k |
| Mid + senior specialist | Area prep (Tuesday) | 5,400 | R$ 73 | R$ 394k |
| Mid + senior specialist | Loose alignments (Wednesday) | 1,500 | R$ 73 | R$ 110k |
| Mid + senior specialist | Post-committee cascade (Friday) | 3,600 | R$ 73 | R$ 263k |
| Leadership subtotal | the room alone | 1,200 | weighted | R$ 360k |
| Long-tail subtotal | the pyramid below | 10,500 | weighted | R$ 766k |
| Aggregated pool | one committee, complete | 11,700 | weighted | R$ 1.13M/year |
10,500 person-hours a year to feed one weekly committee. More than double what the leadership room cost on its own (R$ 360k). And that was just one pyramid. The company had four other rituals of the same make, each with its own base underneath, none of them on anyone's tab.
The calculation that stopped the board
The CFO added up the pieces, and each one fits on its own spreadsheet:
- R$ 360k/year of senior payroll parked in the committee room
- R$ 766k/year of middle and specialist payroll in the pyramid of a single committee
- The cost of deciding late: 31% of decisions sat stuck, with 28 days on average between noticing the need and calling it. On the conservative count, R$ 1.2M/year of revenue arriving late.
- Attention sliced across the three layers, each interruption charging its toll: R$ 1.4M/year of focus that evaporated along the hierarchy
- And the same arithmetic repeated across the other four rituals of the same make (product committee, growth committee, business review, quarterly area review): around R$ 2.1M/year more, in middle and specialist payroll alone
Total recoverable, on the cautious count: R$ 1.84M/year in leadership plus R$ 4.3M/year in the long tail. About R$ 6.1M/year. That is 25 to 28 senior hires the company did not need to make, or 7% to 9% of the entire white-collar payroll, burning on the coordination nobody measured.
The story shifted its axis. The Thursday committee was the visible part, the one that caught the eye. The real money was scattered across the pyramid, exactly where nobody was looking.
The fix was not killing the committee
The first temptation, kill the committee, fell apart in half an hour. The ritual had a legitimate job: align strategy, raise what crosses areas, formalize the collective decision that needs people in the room. The problem was never that the committee existed. It was the whole pyramid it held up underneath, without anyone summing its weight.
Each craft in its place. The economic read is one thing: split the payroll by area, by level, by ritual and by decision, with an auditable number attached to every line. Deciding to act is another, and that fell to the COO and the CFO, in the forum where capital gets discussed. Running the redesign is a third: align who needs aligning, rebuild the document templates, train leadership and management in the async format, log decisions in an indexed registry, track adoption for 90 days. That work belonged to the chief of staff with the area heads. Three distinct crafts, and confusing them is what keeps the number from ever showing up.
With the mandate from the COO and the CFO, the redesign moved in four steps:
- Separate the cadences by real dependency. Operational decisions left the room and started circulating as an async report every Wednesday at 6 PM. The 27% that genuinely called for a live conversation fed a monthly 90-minute committee with 8 people, not 12. Same ruler on the prep meetings: 4 of the 12 went async, 5 became biweekly with pre-reading, 3 survived with half the people.
- Mandatory pre-reading, top to bottom. Every item went into a document closed 24 hours ahead. Whoever showed up without reading lost their voice. Average discussion time per item dropped from 18 minutes to 7 in the leadership committee, and from 12 to 4 in the area meetings.
- Decision log indexed by level. Time-to-decide stopped being a perception and became a measurement. Within 60 days, the average fell from 28 days to 17 in leadership, and from 14 to 9 in middle management.
- End of the staged update, on every floor. 42% of decisions reached the committee already made, just to go on the record. They left the scope entirely. The same rule went down to the area meetings: a direct word between whoever decides and whoever needs to know, with no stage in between.
The number 90 days later
| What moved | Before | After | Equivalent |
|---|---|---|---|
| Time-to-decide, leadership | 28 days | 17 days | -38% |
| Time-to-decide, middle management | 14 days | 9 days | -35% |
| Senior payroll recovered | zero measured | R$ 1.8M/year | ~7 seniors |
| Long-tail payroll recovered | zero measured | R$ 4.3M/year | ~28 in the long tail |
| Aggregated pool recovered | zero measured | R$ 6.1M/year | ~35 people not missing |
| Operating leverage | 0.93x | 1.15x | the long tail pulled more |
The headcount request for the next quarter fell by half. The CEO took the board a number that defends itself, not a hunch to be debated.
The CFO summed it up in the capital committee: "this is the kind of number that dilutes CAC instead of inflating it. And it confirmed the suspicion I could not prove, that we had been looking at the wrong floor the whole time."
What this teaches your company
If your company has a few hundred people, has passed Series B and the board is pressing on the Rule of 40, there is a real chance that a meaningful slice of your white-collar payroll is stuck in inertia rituals. Leadership is the sample that catches the eye, with up to a quarter of the calendar in rituals with no decision on the other side. The long tail almost always sums to more, because it is far more people: the weight is the headcount, times the hourly cost, times the share of time.
The questions that matter:
- Which rituals were designed for today's problem, and which are leftovers from when the company was a third of the size?
- Where does the real decision live and where is it just the relay of a decision already made, on each floor of the pyramid?
- Which pyramid does each leadership ritual fire downward, and how much does it weigh in the aggregate?
- What is the number that defends itself, in reais per year and in days-to-decide, cut by level, that your board can see at the next quarterly review?
That number does not come from an engagement survey or a tool that transcribes meetings. It comes from the economic read of coordination: the payroll stuck in ritual, cut by area, by ritual and by level, with a real cost attached to every line, with no individual name anywhere.
Have you measured your aggregated pool?
The Coordination Calculator estimates the order of magnitude in 30 seconds, with three inputs, triangulated against auditable public benchmarks.
Open calculatorSources: Microsoft Work Trend Index 2023 · Atlassian State of Teams 2024 (n=5,000) · Harvard Business Review (Perlow et al., 2017) + Rogelberg (2019) · Robert Half Brazil Salary Guide 2025