Ask most people how their year is going and they will tell you what they billed. It is the number that is easy to find, it feels like the score, and it is the one your invoicing tells you without being asked.
It is also the number least likely to answer the question you actually care about, which is whether the work is worth doing. Revenue says how much passed through. It says nothing about how much stayed, and nothing at all about which work was worth having.
Why do revenue and profit diverge?
Because the costs of a job are not proportional to its price, and they are not evenly distributed across your clients.
Two jobs bill 5,000 dollars each. One is work you have done before, for a client who briefs clearly and approves in a day. The other is a client who changed direction twice, needed four meetings you did not quote for, and paid sixty days late. Same revenue. Nothing like the same profit.
The gap comes from four places, and only the first is obvious.
Direct costs. Anything you bought for the job. Subcontractors, stock imagery, materials, a licence you needed for that client alone.
Your time, at a real rate. Not just the hours you billed. The hours you spent. Scoping, revisions, meetings, the email thread, the invoice, the chasing of the invoice.
Cost of delay. A client who pays at sixty days is borrowing from you for a month at no interest. That has a cost even when the money eventually arrives, and it is a real one if you have ever moved a purchase because of cash flow.
Opportunity cost. The job you could not take because this one ran long. Hard to quantify, easy to feel, and the thing most likely to be the actual answer to why a good year felt bad.
What counts as a cost on a job?
Anything that would not have happened if you had not taken the job.
That test does most of the work. Your accounting software subscription is not a job cost, it is a cost of being in business, and spreading it across jobs tells you nothing useful. The stock photo you licensed for one client’s brochure is a job cost. The subcontractor you brought in for one build is a job cost. The three hour round trip to their office is a job cost even though nobody invoiced for it.
The category people miss is unbilled time, because it does not appear anywhere. Nothing prompts you to record the hour spent rewriting a proposal, so it never lands against the job, so the job looks more profitable than it was. If you only track one new thing, track that.
Job costs also need to stay off the client’s document. What you paid a subcontractor, and what you added on top, is your business rather than theirs. Ledger keeps private job expenses separate from the invoice for that reason: the cost sits against the job, the client sees only what they agreed to pay, and the margin is visible to you without ever being visible to them.
How do you track cost per line?
Per line item, not per job, if you can manage it.
Job level margin tells you a job made money. Line level margin tells you which part of it did, and that is the actionable version. Most people find that their work splits into things with good margin and things they do because clients ask, and those two are mixed together inside jobs that look fine in total.
A worked example. A design job bills 6,000 dollars: 3,500 for the design work, 1,500 for a print run you arranged, 1,000 for revisions. The print run cost you 1,250, so it made 250 dollars for a fortnight of coordination and the risk of the printer being late. The revisions took three times the hours you assumed. The design work was excellent. In total the job looks fine, so nothing prompts you to notice that you are running a print brokerage at a four percent margin as a favour.
You cannot see that at job level. You can see it immediately at line level, and once you have seen it three times you will either price print properly or stop offering it.
If your tooling will not do this, a spreadsheet with job, line, price and cost will. The discipline matters far more than the software.
Which jobs quietly lose money?
There are patterns, and they repeat across trades.
Small jobs with full overheads. Every job carries a fixed administrative load: quoting, setting up, invoicing, chasing. That load is roughly constant whether the job is 400 dollars or 4,000. Below some threshold, the overhead eats the job. Most people’s threshold is higher than they think, and the fix is usually a minimum engagement rather than a higher rate.
Fixed price work you did not scope. Covered in quote, estimate or invoice. A fixed price against an undefined scope is a bet, and you are not the favourite.
Work outside your normal line. The thing you said yes to because they asked. It takes longer because you have not done it before, it is priced by guesswork, and it rarely repeats often enough to get efficient at.
Clients who pay late, every time. Their cost is invisible in the job and real in your year, and it compounds with everything in getting clients to pay on time.
Your favourite client. Said carefully, because it is common. The client you like, who has been with you for years, on rates set years ago, who now gets more of your attention than they pay for. Long relationships drift towards the client’s advantage unless somebody reprices them, and that somebody is you.
Am I charging enough?
Turn the question into one that has an answer: what did you actually earn per hour worked, across a real period, after job costs?
Take a quarter. Add up what you invoiced. Subtract direct job costs. Divide by every hour you spent on client work, billable or not, including scoping, revisions, meetings and admin attached to jobs. That number is your effective rate, and for most people it is meaningfully below their nominal rate, because the nominal rate is applied only to billable hours and the effective one is diluted by all the others.
Then compare it to what you need. Not what you would like: what covers your costs, your unpaid time, your holidays, your tax, and a margin for the months that go quiet.
If the effective rate is below the number you need, more work will not fix it. That is the important part. Taking on more of the same work at the same prices scales the problem, and the busiest year of your life can be the one where you end up further behind.
How do you use it to reprice?
Repricing everything at once is how people lose clients. Do it in order.
Start with new clients. Your new rate applies to everyone who has not worked with you before. There is no conversation to have and no history to manage. Within a year, a meaningful share of your work is at the new price without you having negotiated once.
Stop the loss-makers. For the categories that consistently lose money, either price them at what they cost, or stop offering them. A four percent margin print brokerage is a hobby with paperwork.
Raise the drifted ones next. Long standing clients on old rates get a straightforward, unapologetic note with a date on it. Most stay. Some do not, and the ones who leave over a fair increase were the ones eating the margin.
Set a floor. A minimum engagement size, below which the overhead makes the job not worth doing. This is easier to hold than a rate, because it is a rule rather than a judgement.
None of this requires new software. It requires cost recorded against the work it belongs to, at the time, rather than reconstructed at year end when you have forgotten which job the subcontractor invoice was for. Whatever tool you use, that is the one habit worth building.
Revenue tells you the business is busy. Margin tells you which parts of it are worth being busy on. They are different questions, and only one of them changes what you do on Monday.