Learn

Your books are a spreadsheet and a folder of receipts. Here's what that quietly costs you.

Nearly every early founder keeps their books the same way: one hand-built spreadsheet, plus an inbox of receipts. It feels like enough. It structurally isn't, and here is why.

By Pelle Brændgaard

I have kept a company’s books in a spreadsheet. One tab that was, in the perfect words of one founder, “a crazy cross between an income statement and a balance sheet that seems to switch back and forth every line.” It felt like being on top of things. It was not. Even later, at the company I run now, the outsourced service we used only closed the books once a quarter, so between those closes we were flying on a spreadsheet regardless.

Here is the core problem, and it took me embarrassingly long to internalize.

A spreadsheet records cash. Books record what you earned.

Your bank balance tells you what happened to your account. Real books tell you what your business earned and owes, which are different numbers, often wildly different.

Work a real example. In January, a customer pays you $12,000 for an annual plan. Your spreadsheet shows $12,000 of cash and a great month. But you have not earned $12,000. You have earned one month of it, $1,000, and you are holding the other $11,000 as something closer to a debt: if they cancel in March, you may owe most of it back. On real books, that $11,000 sits as deferred revenue, a liability, not profit.

A founder summed up the whole gap in one line:

“You care about cash; investors and IRS cares about accruals.” (YC founder on private YC forum)

The spreadsheet only ever shows you the cash half. So it flatters you in the months you collect and terrifies you in the months you don’t, and neither picture is your actual business.

What is a spreadsheet structurally unable to tell you?

This is not about working harder in Excel. A spreadsheet has no concept of double-entry, which is the two-sided system real accounting runs on: every dollar comes from somewhere and goes somewhere, and the two sides must balance. Without that, three questions have no reliable answer:

What did I actually earn this month (not collect)?
What is my margin on this kind of customer, once you count the real cost of serving them?
What do I owe that has not hit my bank account yet: the deferred revenue, the unpaid bills, the taxes?

You can hand-fake any one of these for a board meeting. You cannot keep them all correct and consistent by hand, which is why founders describe manual books as perpetually stale:

“The biggest frustration with manual accounting is you don’t get your books until weeks after the month ends, so any budgeting or forecasting is already one to two months out of date.” (YC founder on private YC forum)

Doesn’t this fix itself with QuickBooks?

That is the reasonable next move, and for most founders it stalls fast. YC’s own CFO is blunt that doing it yourself badly “can cause some brain damage and result in expensive mistakes.” But doing it in software built for a bookkeeper is its own wall, which is a whole separate pain I wrote about in “I did the trial and immediately logged out.”

You do not need to become an accountant

Here is the honest part. The answer is not “learn double-entry accounting,” and it is probably not “hire a bookkeeper” yet either. It is that the books should be kept correctly underneath you, by something that already understands double-entry, so you can just ask them what you earned. What that something is has changed recently, which is the thread running through everything else here.

First, though, one specific way the spreadsheet breaks that catches more founders every month: getting paid in stablecoins.