What a bookkeeper should have finished in your first ninety days
Hiring a bookkeeper is one of the few purchases where you cannot tell whether you got what you paid for until months later. A bad hire does not announce itself. The invoices still go out, the bank feed still updates, and the reports still arrive on time. What you actually bought is a set of judgements about your accounts, and those take a while to show up as either right or wrong.
So here is what I think should be true at thirty, sixty, and ninety days. Not because there is one correct way to run a bookkeeping engagement, but because a bookkeeper who cannot hit these markers is probably not going to, no matter how long you wait.
Days one to thirty: cleanup and access
By day thirty, the bank feed should be complete and reconciled through at least the prior month. Not partially. Not mostly. Complete. Every account connected, every statement period closed, and a reconciliation report you can open that shows the ending balance matching the bank. If you are three weeks in and there are still two accounts not connected because someone is waiting on credentials, that is a warning. Connecting accounts takes an afternoon. Waiting on it usually means nobody owns the task.
By day thirty you should also have a written list of what was wrong when they started. This is the single most useful document in the first month and almost nobody produces it unprompted. It does not need to be long. Six lines is fine. Uncategorized transactions from last year, a duplicate vendor, sales tax not being tracked separately, a loan recorded as an expense. What you want is evidence that someone actually looked, rather than moved transactions into categories until the screen went green.
Days thirty to sixty: the close rhythm
By day sixty, the chart of accounts should be stable and it should match how you run the business. This is the part founders underestimate. A generic chart of accounts will produce a profit and loss statement that is technically correct and practically useless, because it groups your revenue in one line and your costs into categories that do not map to any decision you make. If you price by plan tier, you want revenue split by plan tier. If you have contractor costs that behave differently from salary, they should not share a line. By day sixty you should be able to look at the statement and answer a real question without exporting to a spreadsheet.
By day sixty you should also have a repeatable close. Not a heroic one. A close with a checklist, a target date, and a person who owns each step. Ask your bookkeeper what their close process is. If the answer is a description of their personal work habits rather than a sequence of steps with a date attached, the close will slip the first month they get busy, and it will slip quietly.
Days sixty to ninety: reporting you can use
By day ninety, you should have three months of statements you trust enough to make decisions from. That means the numbers did not get restated after the fact, the categories did not change halfway through, and the reports arrived on the same date each month. Consistency matters more than sophistication at this stage. A simple set of accounts produced the same way three months running tells you more than an elaborate one that gets rebuilt every time.
By day ninety you should also know what your bookkeeper needs from you and how often. The good ones are specific about this. Receipts for anything over a threshold, a note when you take a draw, a heads up before you sign a new software contract. If by three months you still do not know what they want, they have not told you, and you are going to be the reason the books are late.
Now the thing that means you hired wrong. It is not a missed deadline. Anyone can miss a deadline once. It is a bookkeeper who cannot explain a number. Pick any line on your profit and loss statement and ask why it moved. A good one answers in one sentence, in plain English, without opening the file. A bad one says they will look into it and gets back to you with a screenshot. The screenshot is the tell. It means they are reading the software output for the first time too.
One more marker, and it is about you rather than them. By ninety days you should be asking better questions than you were at the start. If the reporting has not changed what you pay attention to, you are paying for compliance rather than insight. That is a fine thing to buy, but you should know which one you bought.
If you are ninety days into a bookkeeping arrangement and you cannot say whether it is working, that is a fair reason to get a second read. We will go through three months of your accounts and tell you plainly what is solid and what is not, for a flat one-time fee, no retainer required. You can start with a free fifteen minute Books Check here: https://app.usemotion.com/meet/GainesCapitalBooks/GCBDiscover
Matt
Matt runs Gaines Capital Books, a flat-rate bookkeeping and CFO shop for SaaS founders.
