A lot of owner-operated businesses are quietly two businesses wearing one name. A studio that shoots weddings on weekends and corporate work during the week. A shop with a retail counter and a wholesale account list. A consultant with project work and a couple of retainers. One bank account, one set of books, one number at the bottom.
That number is usually fine. It is also usually useless, because it is an average of two things that behave nothing alike.
The average hides the answer
Say the business does $200,000 and runs a 40% gross margin. That reads as a healthy year. Split it and you might find one side runs at 55% and the other at 25%, and the low-margin side is the one eating most of your week. Now you have a decision in front of you instead of a summary behind you: raise price on the weak side, stop taking that work, or accept it deliberately because it feeds the other side.
You cannot make that call from one blended number. You cannot even see that the call exists.
An average of two different businesses describes neither of them.
What it takes to actually see it
The good news is that this is a bookkeeping problem, not a software problem. QuickBooks can tag every transaction to a side of the business and produce a profit and loss for each one. What determines whether that is cheap or expensive is not the software. It is how your money moves.
If each side already has its own bank account and card, this is close to free. Every transaction arrives already labelled by where it landed, a rule sorts it on the way in, and the only real work each month is splitting the shared costs: rent, insurance, your phone, the software you both use.
If both sides run through one account, it is a different job. Now something has to decide which side a purchase belongs to, and for a good number of transactions the bank feed simply cannot tell. Somebody has to know, and that somebody is usually you, once a month, answering questions about charges from three weeks ago.
Which is why the single most useful thing most two-sided businesses can do costs nothing: open a second account and card, and put one side on each. Twenty minutes at the bank buys you a permanently cleaner picture and a cheaper close.
The shared costs are a decision, not a formula
Rent does not belong to one side. Neither does your own time. Somebody has to pick a method for splitting them, write it down, and stay consistent, because the method changes the answer. Splitting shared costs by revenue makes a high-revenue, low-effort line look worse than it is. Splitting by hours does the opposite.
There is no correct answer, only a defensible one. Pick the method that matches how the costs are actually driven, put it in writing at the start, and do not change it mid-year because you did not like a result.
When it is not worth doing
If one side is under about 10% of revenue, leave it alone. You will spend more attention on the split than the answer is worth. The same is true if the two sides share every cost and every hour, because then almost everything lands in the shared bucket and the split becomes an opinion rather than a measurement.
And if the two sides are separate legal companies with separate tax filings, this is not the same thing at all. That is two sets of books, and pretending otherwise creates a mess your CPA has to unwind later.
The short version
If your business has two distinct kinds of work and you have never seen them apart, you are probably subsidising one with the other and calling it a good year. Finding out costs less than you would think, especially if you separate the money at the source. It is one of the few bookkeeping decisions that changes what you do on Monday.