Greg Hoy
Timesheets. Just typing the word gives me the same feeling as a line at the DMV. Nobody likes filling one out: not the designer, not the developer, not the project manager who has to badger everyone on Friday afternoon, and not the agency owner, who quietly stopped logging his own time back in year three.
So when a smart agency founder stands up on stage and declares the whole practice a relic, tracing it back to a guy with a stopwatch on an 1878 steel mill floor and an ad man in 1962 who wanted to win a pitch, the room nods along (old gold). I ran Owner Camps for years, and a version of that argument came up in almost every session. I nodded too, because it gets the core problem right: selling time is a clumsy way to sell knowledge work. It rewards the slow, punishes the quick, and turns the agency’s attention inward on how long things take instead of outward on whether they worked. Fixed fees are better for anything you can scope. I still bill hourly for open-ended advisory work, because there’s nothing to scope, and that works fine too. How you invoice is a choice.
Here’s where it gets messy. Somewhere in the last decade, “don’t sell time” morphed into “don’t track time.” Those are two different sentences. One is a pricing decision. The other is driving with your eyes closed.
The timesheet was innocent first
The funny thing is, the history the anti-tracking crowd cites proves this point for them. When the first law firm introduced timesheets in 1945, it wasn’t to bill clients by the minute. They wanted to know what cases were costing them. The billable hour didn’t show up until fifteen years later, when bar associations noticed that firms billing by time made more money. An internal tape measure got turned into a price tag. The fix is to stop using it as a price tag, not to throw out the tape measure.
Same with the steel mill. The market set the price of steel, not the mill’s cost. That’s a decent argument for value pricing. But the mill still knew its cost to the penny, because if a ton of steel costs more to make than the market pays, you find out fast.
What the income statement can’t tell you
Ask an owner how they know what things cost without logging hours and the answer is usually the income statement. Revenue at the top, payroll and rent in the middle, profit at the bottom. Easy.
Except the income statement is like checking your bank balance at the end of a vacation. It tells you how much you spent, not which dinners did it. It won’t tell you that the $60,000 fixed-fee project you estimated at 400 hours ate 520, dropping the margin from 53 percent to 39. It won’t show you which client is quietly running on infinite revisions, or that your senior strategist spent half the month doing coordinator work at a coordinator rate. All it sees is that profit went down while payroll stayed the same. The why is somebody else’s problem.
Every number that answers the why, project margin, effective rate, realization, is hours times a cost rate. The income statement has neither, and it reports in April, long after there was anything to do about it.
Fixed fees need hours more, not less
This catches people off guard. Hourly shops catch a blown budget faster, because the client starts complaining about invoices mid-project. Uncomfortable, sure. But you’re fixing the leak in week six, while there’s still a project to save.
A fixed-fee agency with no time tracking doesn’t find out until tax season. The project feels like a win because the check cleared. Meanwhile 120 ghost hours evaporated, which is the classic agency curse: a busy year with no extra profit to show for it, and a team sending out the next round of proposals built on the same broken estimates.
Value pricing runs into the same trap. Charging on impact instead of hours makes sense; a $40,000 brand refresh for a high-growth startup should be priced on what it’s worth to them, not on the clock. But value pricing only tells you what to invoice. It says nothing about whether you made a dime. If that $40,000 job takes 500 hours because of endless committee feedback, your senior team just worked for $80 an hour and nobody noticed. Value pricing did its job. You took the instrument off the other side of the scale.
Small, senior-heavy shops can hide these leaks for a long time, because the partners absorb the pain. Three co-founders can swallow 500 extra hours and keep going. A twenty-person team with a payroll to make can’t, because the price was set on value and the cost was set on headcount, and nobody is watching the second number.
I ran my own shop this way for twelve years. Here’s what one of those years looked like.
From 2006 to 2018 I managed Happy Cog with a spreadsheet that would make an accountant pull out a monocle. Every pay period got its own tab for each person, every logged hour grouped by client and project. Paycheck divided by 80 gave an hourly cost. Billable hours times our rack rate gave revenue. The difference was a profit or loss per person, per pay period, rolled up every quarter.
Pull up the 2010 workbook and it’s twelve people, twenty-four pay periods, and about 22,000 logged hours. Twelve percent of them, roughly 2,700 hours, are logged to a task called “Miscellaneous.” Not a project. Not even a category. A shrug with a timestamp. Utilization ran over 100 percent in more than half the person-periods, meaning people were logging more than 80 hours in two weeks, which is a burnout problem or a fiction problem and both are bad. And a party site for a conference, built for fun, swallowed 838 hours, which at our rate was a small client’s annual budget.
None of that showed up on the income statement, which just said revenue was decent and payroll was high. It had no way of telling me that a third of our working hours had no client attached, or that a chunk of those had no name. The spreadsheet did, and each problem got its own fix: a real project code for the party, a conversation about the miscellaneous bucket, and a look at who was logging fourteen-hour days and why.
Early versions also had a leaderboard column ranking everyone by profitability. That’s the stopwatch pointed at the person, the exact reason people hate timesheets, and I built it into the header row.
The spreadsheet had another flaw, and it’s the one this article is really about. It assumed every logged hour got billed at full rack rate. It ignored discounts, write-offs, and fixed-fee overruns, so the profit column always looked sunnier than the bank. Billability told me people were working. It never told me whether the work counted. That gap is called realization, and it took me longer than it should have to start measuring it.
Timesheets are fiction. Agreed. Now fix that.
The strongest argument against tracking is that self-reported time is basically fan fiction. Rebuilding your week at 4:30 on a Friday is guesswork and wishful thinking, and 2,700 hours of “Miscellaneous” says it was true of my data too. Bad data in, bad decisions out. If that’s the choice, no data is honestly better.
It isn’t the choice. You don’t need precision. You need low friction:
- Log it daily, not weekly. A week reconstructed from memory is a story. A day is a receipt.
- Track the project, not the task. “Client X website, 4 hours” is all the cost data you need. Six-minute increments are for lawyers, and they hate them too.
- Half hours and whole hours. Precision you don’t have is fiction with more decimals.
- Use it to fix the next estimate, not to police the last one. Nobody looks at an individual’s number except to figure out why the scope was wrong. I say that as someone who ranked twelve people by it and got exactly the culture you’d expect.
What you measure
What you measure is what you end up valuing. Obsess over individual hours and you get padded timesheets and zero trust. I have the spreadsheet to prove it.
Point the same instrument at the estimate instead and everything changes. You get a culture that cares whether the scope was right, which is the thing that protects margins, protects weekends, and makes you better at asking for a real number in the first place.
Price on value where you can. Pick whatever billing model fits. And at the end of the day, take four minutes and write down where the hours went, because someday you’re going to want to know what it cost.