What an Empty COO Seat Costs You Per Month

Most founders treat the COO question as a spending decision: can we afford the seat yet. That framing hides the real question, because the empty seat is already costing you money every month. The cost is simply distributed somewhere less visible than a payroll line.

Below is how we price that cost inside client businesses. A four line ledger, a worked example you can run against your own numbers, and three things to do on Monday. Every figure in the example is a stated assumption rather than an observation, and it is labelled as one.

What is an empty COO seat actually costing you?

It costs you in four places, every month, and none of them sends you an invoice. An unfilled operating seat does not reduce the work that seat would have done. It redistributes that work to people and places where nobody is counting it.

The empty COO seat is not a line you are not paying. It is a line you are paying in four other places.

There are only four places the work can go. It lands on your hours. It lands on your team's hours, as waiting. It lands on work that never starts at all. Or it lands on cash, as the running cost of trying to fill the seat. We call those four lines the Empty Seat Ledger, and together they are exhaustive. There is no fifth place for the work of an operating lead to hide.

Two of those lines you can price this month, with numbers you already have. Two of them you cannot price honestly, and we do not try. A ledger that puts a number against everything is a sales document, not a ledger.

It is worth being clear that the seat is not decorative overhead. In the largest cross country study of management practice, a one standard deviation increase in management was associated with an increase in total factor productivity of 15% (Bloom, Lemos, Sadun, Scur and Van Reenen, 2014). Management quality behaves like a productivity input, not like a perk. Leaving that input unowned has a price, and the price is not zero.

Why does the cost stay invisible?

Because every line on the ledger is paid in time you have already bought, or in work that never started. Neither of those appears in your accounts. Three mechanics do the hiding, and a fourth keeps the decision deferred.

First, waiting does not change the wage bill. Your team is on payroll whether they are working or waiting for an answer. What changes is what the wage bill buys. A month in which a third of senior capacity sits in a queue costs exactly the same as a month in which it does not.

Second, your own displaced hours look like commitment rather than substitution. Chairing the operations review, unblocking the approval, chasing the supplier: it all feels like leadership. Some of it is. Most of it is you doing a job you have not filled.

Third, the largest line carries no evidence. Deferred capacity is work that never started, so there is no slipped date, no complaint and no artefact to point at. Nobody escalates a project that was never scoped.

The fourth reason is economic rather than psychological. Manpower cost is the loudest number in the room. In the Singapore Business Federation's National Business Survey 2025, Manpower and Wages Edition, rising manpower cost was the top challenge named by 63% of 555 businesses, 85% of them SMEs, surveyed between 25 June and 18 July 2025. A founder comparing a known salary against an unknown drag will defer every time. The comparison is not wrong. It is missing a number on one side.

How do you price it? The Empty Seat Ledger

Price each line against something you can count this month, not against a benchmark. Benchmarks tell you what other companies look like. They do not tell you what your queue costs.

The ledger below is ordered by how visible each line is, not by how large it is. That ordering matters, because the first line is the one founders always name and it is almost always the smallest on the page.

Exhibit 1: The Empty Seat Ledger

Ledger line	What it is	Where you see it this month	How to price it
1. Founder displacement	Your hours spent on operating work the seat would own	Your calendar: approvals, scheduling, chasing, coordination	Transferable hours a week x 46 weeks x a senior day rate
2. Decision latency	Work that waits because the decision has nowhere to go but up	Your inbox: the same four or five questions every week	Decisions a week x days waiting x people blocked x day rate
3. Deferred capacity	Work that never starts, because nobody owns starting it	The manager never onboarded, the process never built	Do not price it. Name it, and accept that the ledger understates
4. Search carry	The ledger keeps running while you look for a full time hire	Months three to six of a search that has not closed	Monthly ledger total x expected months of search

Illustrative structure. The formulas run on your numbers, not ours.

Line one is the line founders quote, and the smallest one here. Line two is usually several times larger, because it multiplies across everyone who was waiting. Line three is larger still, and we refuse to put a number on it. Line four is the one nobody counts at all, and it is the reason a search is not a fix.

Run the ledger every quarter rather than once. The point of a monthly figure is that it moves. A line that shrinks after you hand something over is telling you the handover worked. A line that grows while revenue grows is telling you the seat is now overdue rather than premature.

What does the ledger add up to in a real business?

In the illustrative firm below, roughly USD 8,700 a month, or 10.9% of operating profit, from the two lines we are willing to price.

Take a Melbourne B2B services firm: 44 people, USD 8M revenue, USD 960,000 operating profit, founder led, with no operating lead between the founder and the department heads. This is an illustrative profile, not a client, and every input below is an assumption you should replace with your own.

The assumptions. Four decisions a week reach the founder that an operating lead would settle. The median wait on one of them is three working days. On average 1.5 people are blocked while it waits. Only one third of blocked time is genuinely lost; the rest gets absorbed into other work. Fully loaded cost is USD 330 per person day. The year holds 46 working weeks. Separately, the founder spends 12 hours a week on operating work, and half of that is genuinely transferable to someone else. Founder hours are priced at USD 380 a day rather than at revenue per hour, because pricing them at revenue would flatter the model.

Decision latency. Four decisions x three days x 1.5 people is 18 person days of waiting a week. One third of that is lost, so 6 person days. Over 46 weeks that is 276 person days. At USD 330 a day, USD 91,080 a year, or USD 7,590 a month.

Founder displacement. Twelve hours a week, half of it transferable, is six hours, or 0.75 of a day. Over 46 weeks that is 34.5 days. At USD 380 a day, USD 13,110 a year, or USD 1,093 a month.

Ledger total, two lines priced: USD 104,190 a year. USD 8,683 a month. 10.9% of operating profit. Illustrative, using the assumptions stated above.

Halve the decision count and halve the wait, and the ledger still reads about USD 2,200 a month. The number is not fragile in the direction that matters.

Two further costs sit outside the model entirely, and we have not tried to price either. The first is the decisions that were never escalated at all, because the team learned the queue was slow and quietly took the cheaper option instead. The second is turnover among capable senior people who spent a year waiting for answers. Both are real. Neither is defensible as a number, so neither is in the total.

Two lines are missing from that total on purpose. Deferred capacity is real, and in most businesses we see it is larger than both priced lines combined, but we will not invent a figure for it. Search carry is not a cost of the gap so much as a cost of the usual response to the gap, which is where the arithmetic gets uncomfortable.

Because a search is not quick. In Australia, the recruitment difficulty rate was 49% in the June quarter 2025, measured by Jobs and Skills Australia in a telephone survey of at least 800 employers a month. In Singapore, the share of PMET vacancies unfilled for at least six months rose from 14.4% in 2024 to 16.0% in 2025, after three consecutive years of improvement (Ministry of Manpower, Report: Job Vacancies 2025). The search itself also costs money: SHRM's benchmarking put executive cost per hire at a median of USD 8,750 and an average of USD 28,329 across 212 organisations, from data collected between April and November 2021. That last figure is US member data and should be read as an order of magnitude, not as an APAC quote.

Run a four month search in this firm and the ledger does not pause while you look. Four months at USD 8,683 is USD 34,732 accrued before anyone starts, on top of the cost of the search. A full time hire is the right answer in plenty of businesses. It is simply not a fix for the months it takes to land one.

What do you do on Monday?

Three things, none of which require hiring anyone.

One. Count the queue, not your hours. For one week, log every decision that reached you that somebody else could have made: the date it arrived, the date you settled it, and how many people were waiting on it. Then run the multiplication above with your own numbers. Most founders are surprised by the people blocked figure, not by the days waiting figure.

Two. Put both monthly numbers on the same page. The monthly cost of the seat you are considering, and the monthly cost of the ledger you are already running. Compare monthly against monthly. The COO decision feels hard mainly because it is usually framed as an annual salary against a vague feeling.

Three. Size the role to the ledger, not to the org chart. If two lines carry the cost, you need those two lines owned. That is a scope, and a scope can be bought at one or two days a week, reviewed monthly, and widened when the ledger says so. Company size decides the scope of the role. The bottleneck decides whether you need it at all.

One honest caveat from our side. An operator running those two lines drives the execution and owns how the work is run. We do not warrant the business outcome, and anyone who tells you they do is selling you something other than operations.

If you want a second pair of eyes on your own ledger, that is a 30 minute conversation rather than a proposal. You can see how we work at rem-up.com, or book a 30 minute call and bring one week of your decision queue.

Innovate. Optimize. Grow.