There’s a version of me that will spend a Saturday afternoon with a year of one expense category laid out side by side, and genuinely enjoy it.
I want to say that plainly at the start, because most writing about small business finances assumes the owner avoids the books out of dread or incompetence. That isn’t my situation and I doubt it’s yours. I’m good at this part. Give me a stretch of history and a question and I’ll find the thing in it that doesn’t belong. I’ve done it across eight businesses, enough times to have a method.
And I still stopped doing it.
That’s the part worth writing about. Not the skill. The stopping.
What the digging actually is
It isn’t reading a statement. Reading a statement tells you nothing — everything on it already happened and every line on it looks reasonable in isolation. That’s the whole trouble.
The real work is reading one statement against eleven others.
You take a single vendor and follow every payment they’ve ever taken from you, all the way back. Not the amounts on their own — the spacing, the rhythm, the shape. When did this become monthly? Was there a month it changed level and stayed there? Does the gap between charges make sense for what this thing is?
Then you take a category and lay the months side by side, and you look for the step. Not the spike — the step. The spike gets found by anybody. The step is a number that quietly went up and never came back down, and it’s only visible when you can see enough months at once for the line to have a shape.
Then you take one card and ask which entity should have paid for each charge — a different question than whether it was recorded correctly. It usually was.
Then you go looking for the same thing wearing two different names.
None of that is accounting, exactly. It’s closer to reading a long transcript for the sentence that contradicts something said forty pages earlier. It’s basically financial forensics — and there’s a satisfaction in it that has nothing to do with saving money. You’re looking for the one thing that doesn’t fit, and when you find it, you found it.
The last real one I pulled on started as an absence, not a number. A charge — over a thousand dollars — with no receipt behind it.
That’s the tell. Not a figure that looks wrong, a figure with nothing underneath it. So you start the way you always start: where’s the receipt. There isn’t one. You ask the owner, and the owner has never heard of it. That’s the moment the temperature changes, because now you’re not reconciling anymore, you’re investigating.
So you do the work. You go back through the emails for anything that mentions it. You read the statement itself for whatever it will tell you about who charged it. You look the vendor up. And in this case, at the end of all of it, the honest answer was: we cannot verify this. We don’t know who they are. Which means the thing I’m actually dealing with now isn’t a bookkeeping discrepancy. It’s fraud.
I want to be precise about how that surfaced, because it’s the whole argument for doing this at all: it only became visible through the missing receipt. Nobody flagged it. The bank didn’t stop it. It sat in a feed that balanced. The single thread that led to it was one charge that couldn’t produce its paperwork — and pulling that thread is exactly the work nobody has time to do.
Why I stopped, honestly
If you’d asked me at the time, I’d have told you I was busy. That’s true and it isn’t the real answer.
The real answer is that I didn’t want to find anything.
Sit with that for a second, because it’s not the same as not wanting to do the work. I like the work. What I didn’t want was the outcome. Because finding something means I now own it.
If I find it, I have to email the vendor, dig up whatever we originally agreed to, decide whether it’s worth disputing, and then dispute it — which is never one message. I have to move entries between periods and tell my accountant why. I might have to cancel something and then find out what quietly depended on it. And underneath all of it, sit with the fact that it was happening for a while and I didn’t see it.
The finding takes five minutes. The consequence takes a fortnight of small errands, in a week that had no room before I opened the drawer.
So somewhere below the level of a decision you’d notice making, you close the tab. Not out of laziness. Out of triage. You’re protecting a schedule that’s already over capacity, and the fastest way to protect it is to not learn anything new that needs handling.
If you’ve ever closed your banking tab slightly too fast, that’s what was happening. It’s a rational response to being overloaded. It’s just an expensive one.
The task with no finish line
There’s a second thing making it undoable, and it’s structural rather than emotional.
Almost everything else on my list has an end condition. Invoices go out, and then they’re out. Payroll is submitted, and then it’s submitted. Even a month-end close is bounded — a set of questions with a last question, which is exactly why it can shrink to something you fit into a morning.
Forensic review has no last question.
You don’t finish it. You stop. You stop when the afternoon is gone or your attention is, and the difference between that and actually being done is invisible from the inside. There’s no state where you can say: checked, clean, closed. There’s only how far you got.
Which means you can’t size it. And a task you can’t size can’t be scheduled, because scheduling requires knowing what it displaces. So it never gets a slot in a real week. It goes onto the list of things you’ll do when things calm down, and that list is not a queue. It’s a place where things go to become permanent.
Every busy owner has one of these. Financial forensics is almost always on it.
The bet you’re making every time
Here’s the other thing that quietly trains you out of it.
Every time you consider a proper dig, you’re wagering an afternoon against an unknown return. You don’t know if there’s anything in there. That’s the point — if you knew, you wouldn’t need to look.
And most of the time you find nothing much. Which is the good outcome. It’s also the outcome that teaches you not to bother next time — you spent the afternoon, came up with almost nothing, and the lesson your brain takes is that this wasn’t a good use of a Saturday.
So the better your books are, the less rational the digging feels. Right up until the one time it isn’t, and by then the thing you’d have found has had a long run at your account.
That’s the trap. Being capable isn’t enough. Being willing isn’t either. As long as it costs an afternoon, someone with eight businesses and a full week will correctly decide against it, almost every time.
I’ve written elsewhere about why an outside bookkeeper structurally can’t be the one who catches this — they’re not inside the business and were never hired to notice. This is the other half, and the uncomfortable half. Even when the person with all the context is willing, capable, and actually enjoys the work, the work still doesn’t happen. Not because nobody can. Because nobody has an afternoon to spend on a maybe.
What changes when the looking is free
The fix isn’t a better method. I already had a method.
The fix is removing the decision.
When the forensic pass costs nothing and happens continuously, there’s no afternoon to authorize. Nobody has to weigh it or decide it’s worth it this month. It runs whether you were in the mood or not, and — this is the part that matters — it runs on the months you’d absolutely have skipped, which are the months you’re most likely to have missed something, because they’re the busy ones.
Continuous beats capable. That’s the entire argument. A person who’s excellent at this and does it twice a year is worse protection than something mediocre that never stops looking, because the thing that never stops looking finds things while they’re still small, still recent, still disputable, and still something you can remember agreeing to or not.
You also stop carrying the question. There’s a low-grade weight in half-knowing something is probably off somewhere in your numbers and having no plan to find out. It doesn’t feel like much on any given day. You notice it when it lifts.
What Ledger actually does with this
Ledger is the bookkeeping app I built to sit where a monthly bookkeeper used to sit, and the forensic part is the piece I’d least want to give back.
It holds your transaction history and compares against it — not as a report it produces, as the thing it’s constantly doing. A charge that doesn’t match your own history gets surfaced. It keeps a running list of charges that don’t have documentation attached yet, so the gaps show up while you can still remember what they were. And it keeps the month-to-month reconciliation rhythm going, which is what makes any comparison mean anything — you can’t compare against a history that’s four months stale.
And you can ask it a plain question about your own books and get an answer with the actual records shown underneath it. If it doesn’t know, it says so. It doesn’t fill the gap with something that sounds right.
What it won’t do is tell you what any of it means for your taxes. It isn’t an advisor and doesn’t pretend to be one. It surfaces the line and shows why it stood out. The judgment stays with you; anything tax-shaped stays with your accountant. A flag is a request for two seconds of your attention, not a verdict.
I still dig
I want to be clear that I didn’t hand over something I hated.
I still go in there. Some evenings I’ll follow a vendor back through the history just because I want to understand something, and I still enjoy it exactly as much as I used to.
The difference is what I’m carrying when I open it. I’m not going in to find out how bad it is. I’m not braced. I already know nothing large is sitting in there unexamined, because something has been examining it the whole time I wasn’t. So the digging became curiosity instead of a debt I’d been avoiding.
That’s a much better way to spend an evening. And a much better relationship with your own numbers than the one where you’re the only thing standing between your money and everything that quietly wants a piece of it — doing that job in the gaps, from memory, on an afternoon you never actually have.
If you’re capable of this work and have quietly stopped doing it, you’re not the exception. You’re the normal case, and the reason is arithmetic, not character.
Have a look at what Ledger surfaces. Bring a real month. The first pass is always the interesting one.