Last quarter I sat down with my own numbers, the way I make my clients do, and ran a full expense autopsy on my business. I have been doing this for years. I preach it. I am supposed to be the disciplined one.
I found nineteen hundred dollars a year in tools I was paying for and not using. A transcription service I replaced eight months ago and never cancelled. Two project apps for a team of one. A premium tier on software where I was using maybe a tenth of the features. A domain I bought for an idea that never shipped, auto renewing like a loyal little ghost.
None of it was a big decision. That is the whole point. Businesses rarely die from one catastrophic expense. They bleed out slowly, from a hundred small charges nobody is watching anymore, each one too tiny to notice and too automatic to question.
Why the money leaks where you cannot see it
The modern business runs on subscriptions, and subscriptions are engineered to be forgotten. That is not a conspiracy, it is the business model. A charge small enough to ignore, billed on a cycle just long enough that you forget the last one, renewing automatically so no human ever has to decide yes again.
Nineteen dollars here. Forty nine there. A hundred and twenty a month for something a teammate signed up for and left. Individually they are noise. Together they are a salary. The average small business carries hundreds of dollars a month in what I call zombie spend: charges that are technically alive on your statement but dead in terms of value. They shuffle out of your account every month and give you nothing back.
The reason this survives, quarter after quarter, is that nobody owns the review. Everyone assumes the spend was justified when it was approved, and approval was months or years ago. Without a scheduled autopsy, dead spend simply never gets caught. So we schedule one.
Run the numbers on your own stack for a second. If you are carrying five hundred dollars a month in tools, half of which you would struggle to justify out loud, that is three thousand dollars a year of questionable spend. Over five years, with that money instead invested at a modest return, you are looking at close to twenty thousand dollars. The subscriptions feel like the cost of doing business. Some of them genuinely are. The rest are just the cost of not looking.
The quarterly autopsy, step by step
This is a two hour ritual you run four times a year. That is it. Two hours a quarter to find and kill money that would otherwise leak for the life of your business. I have never once run this and not paid for the next decade of the habit in the first sitting.
Pull ninety days. Export the last quarter of transactions from every account and card the business touches. One CSV, everything in it. Do not filter yet. You cannot autopsy what you refuse to look at.
Isolate the recurring. Sort by merchant and flag anything that shows up more than once. Those repeat charges are your subscription surface, and they are where almost all the rot lives.
Run the kill test on each one. Three questions, answered honestly, decide the fate of every recurring charge.
Cancel on the spot. Not later. Later is where dead subscriptions go to keep living. If it fails the test, kill it before you move to the next line.
The kill test is the heart of the whole thing, and it is brutally simple. For every recurring charge, ask:
Have I actually used this in the last thirty days? Not could I, not might I. Did I.
Is there a free or already paid tool that does the same job? Half of your stack overlaps with something you already own.
Does this charge connect to revenue or to a system I actually run? If you cannot trace a line from the expense to money coming in or a process going smoother, it is decoration.
One yes is usually enough to keep a tool. Three no answers is a death sentence. The discipline is in refusing to negotiate with yourself. The instant you hear but I might need it someday, you have found a zombie.
A teardown, line by line
Let me show you the test in motion with a handful of charges from a real stack, lightly disguised, so you can see how fast the decisions actually go.
A forty dollar a month design tool, used twice in ninety days, duplicated by a free tool the team already has. Two no answers and one weak yes. Killed. Four hundred and eighty dollars a year recovered.
A ninety nine dollar a month analytics platform nobody has logged into since the person who set it up left the company. Zero uses, no revenue connection, fully replaceable. Three no answers. Killed. Nearly twelve hundred dollars a year.
A two hundred dollar a month email and automation tool that every sequence and signup form depends on. Used daily, tied directly to revenue, no real substitute. Three yes answers. Kept without hesitation, and it should probably move to the annual plan to shave another fifteen percent.
That is the whole method in three lines. The tool you keep is not the cheap one, it is the one that earns its place. The autopsy is not about spending less for its own sake. It is about making sure every recurring dollar is pulling weight.
Let AI do the sorting
The tedious part of an autopsy is reading ninety days of cryptic merchant names and remembering what each one is. That is exactly the kind of grunt work you should never do by hand again. Drop your exported CSV into Galaxy.ai and ask it to group every transaction by merchant, flag anything that recurs, and estimate the annualized cost of each recurring charge.
In about a minute you get back the one view that actually changes behavior: not what you paid last month, but what each subscription is quietly costing you per year. Seeing that a forgotten twenty nine dollar tool is really a three hundred and forty eight dollar a year commitment is what finally makes you cancel it. The monthly number lets you rationalize. The annual number does not.
Ask it one more thing while you are there: which of these tools overlap. The AI will happily point out that you are paying three separate companies to do roughly the same job, which is the most common and most expensive pattern in any stack over a year old.
The reason the annual view works where the monthly view fails is pure psychology. Your brain prices a subscription at its monthly sticker, because that is what it sees, and nineteen or twenty nine dollars always feels trivial. But you are not renting these tools for a month. You are renting them, on autopilot, for years. Forcing every charge into its true annual and five year cost is the single most powerful reframe in the entire autopsy, and it is the one thing people consistently skip.
Watch for shadow spend
The nastiest zombies are the ones you never personally approved. In any business past a couple of people, teammates sign up for tools on their own cards or through self serve trials that quietly convert to paid. These shadow subscriptions never cross your desk, so they never get questioned, and they can stack into hundreds of dollars a month without anyone deciding they should.
The fix is not surveillance, it is a single shared list. Every recurring tool the business pays for gets one line on one document, with an owner, a cost, and a purpose. If a charge is not on the list, it does not get paid next cycle, full stop. The list turns invisible spend into visible spend, and visible spend is the only spend you can actually manage.
Expect resistance from the tools themselves, too. Companies make leaving harder than joining on purpose. Signing up is one click, while cancelling is a support ticket, a retention offer, and three confirmation screens engineered to wear you down. Budget five minutes for the friction and do not let a well designed cancellation flow guilt you into keeping something you already decided to kill. The friction is not a sign you should stay. It is a sign they know most people will not bother.
Automate the follow through
Finding the leaks is only half the win. The other half is making sure they stay dead and that new ones get caught fast. This is where a little automation turns a one time cleanup into a permanent defense.
I use Make.com to run a standing watch on spend. Two automations do most of the work.
New charge radar. When a new recurring merchant appears that was not on last quarter's list, Make flags it and asks you a single question: keep or kill. New zombies get caught in their first month instead of their first year.
The quarterly nudge. On the first day of each quarter, Make schedules the autopsy on your calendar, pulls your reminder checklist, and will not let the ritual quietly disappear from your routine. The whole system is designed so the review happens even when you forget it exists.
There is a deeper reason to protect your time here, and it is worth saying plainly. Every hour you spend manually policing forty dollar charges is an hour you did not spend on the work that actually grows the business. If you want to see how much your hours are really worth before you decide what to automate, a tracker like Rize.io will show you where your week is going with uncomfortable precision. Once you see the number, the case for automating the boring defense makes itself.
THE AUTOPSY MATH Two hours, four times a year. If your first teardown finds even two hundred dollars a month in dead spend, that is twenty four hundred dollars a year recovered for eight hours of work. Almost nothing else you do pays three hundred dollars an hour. |
The honest caveats
A teardown is a scalpel, not a chainsaw. A few warnings so you do not overcorrect.
Cheap is not the goal, useful is. The point is to cut dead weight, not to starve tools that actually earn their keep. A three hundred dollar tool that saves you ten hours a month is a bargain, not a target.
Watch for the cancel and re buy cycle. If you keep killing and re subscribing to the same tool, it is not dead spend, it is a real need with bad billing. Move it to annual and stop the churn.
Do not autopsy in anger. Run this cold and scheduled, not in a panic during a slow month. Panic cuts the wrong things.
Your move this week
Do not wait for the quarter to turn. Run a mini autopsy right now while the idea is hot.
Export the last ninety days from your main business account and card.
Feed it to an assistant and get back the annualized cost of every recurring charge.
Run the three question kill test and cancel every zombie you find, today, not later.
Schedule the next full autopsy on your calendar before you close this email.
My own teardown last quarter took ninety minutes and recovered nineteen hundred dollars a year. That is not a one time win. That is nineteen hundred dollars that now compounds somewhere useful, every single year, because I spent one quiet morning looking at what I was actually paying for.
AN OFFER FOR YOU ▸ REPLY: AUTOPSY Want my exact teardown system? Reply with the word AUTOPSY and I will send you the Expense Autopsy Kit: the categorization prompt I feed the AI, the kill test scorecard, and the Make automation that catches new zombie charges in their first month. Free, just hit reply. |
Your business is almost certainly bleeding right now from cuts you stopped noticing months ago. The bleeding is quiet, it is small, and it is completely fixable in a single morning. Schedule the autopsy. Find the leaks. Then keep the money that was always yours.
Build the system. Skip the guessing.
Alex
The Wealth Grid | Issue 76 | July 1, 2026
Know someone who is tired of guessing with money? Forward this along. New here? Subscribe and read the archive on Beehiiv.