Why an Agency's Time Tracking Is a Pricing Problem Disguised as an Admin Problem

The timesheet argument you have had every Friday

Somebody in operations sends the reminder. Half the team fills in their week from memory on Monday morning. A project manager spends part of Tuesday chasing the rest. The totals go into a report that nobody fully trusts, and the next quarter someone proposes a new tracker, because the current one clearly is not working.

If that cycle sounds familiar, you have probably also been shown the number that justifies breaking it. Agencies lose 15 to 30 percent of billable time to manual timesheets, and project managers spend 3 to 5 hours a week chasing and reconciling entries. Those figures rank on page one of every search on the subject (Shiftflow, Teamwork).

We do not dispute that the hours go missing. We dispute what the missing hours mean. The standard framing treats them as leakage from a correct price. For most agencies we look at, they are the opposite: evidence that the price was wrong, arriving too late to do anything about it.

Why the leakage math only works for one kind of agency

The leakage story is true for an agency that bills by the hour. If you sell time and fail to record it, you fail to invoice it, and the revenue is gone. Capture more of it and you bill more of it. A tracker pays for itself in that world.

Very few mid-market agencies still live entirely in that world. Most of the book is retainers, fixed-fee projects, and scoped deliverables. In a fixed-fee engagement, the hour is not the unit of sale. The client pays the same amount whether your team logged 40 hours or 90. An unrecorded hour does not cost you a dollar of revenue directly, because there was never an invoice line for it to appear on.

So what does it cost you? The ability to know. The quick website refresh that was quoted at 10 hours and consumed 25 is the example one of the tracking vendors uses itself, and it notes that you discover the overrun after delivery (Productive). That is not an admin failure. That is a scoping and pricing failure, and the timesheet is merely where the evidence was supposed to show up.

Once you see it that way, the recovery pitch falls apart. You cannot "recover" 15 hours of a fixed fee by recording them more accurately. You can only learn, for next time, that this kind of work costs more than you are charging for it.

Why nobody fills the form in

The usual explanation for poor time entry is discipline. People are busy, they forget, creative staff dislike admin. All true, and none of it explains why the problem survives every new tool, every reminder bot, and every policy memo.

Here is the explanation that does. People will not reliably fill in a form whose output is used to audit them.

Watch what happens to timesheet data inside a typical agency. It does not feed a pricing model. It feeds a utilisation report, and the utilisation report is used to ask individuals why their number is low. Producer utilisation below 65 percent gets described as a sign of overstaffing or underpricing (Iota Finance). Notice that the second diagnosis is about the price, and the conversation that follows is almost always about the first, which is about the person.

Staff learn this fast. They log time to the budget they were given, not to the work they did. A 20-hour task goes in as 12 because the job code only had 12 left, and the other 8 get spread across internal meetings or quietly dropped. The data is not missing at random. It is shaded, consistently, in the direction that protects the person entering it, and that direction happens to be the one that hides your underpricing.

A better tracker does not fix this. A tracker that captures time automatically can record where people spent their minutes, but it cannot record which scope those minutes belonged to or why the task took longer than planned. That is the part a pricing decision needs, and it is the part that requires a human who has no reason to shade the answer.

Who is telling you the number

Take one step back and look at the sources for the 15 to 30 percent figure, the 3 to 5 hours of chasing, and the 65 percent utilisation threshold. Every one of them is published by a company that sells time tracking or agency management software.

That does not make the figures false. It does tell you which question they were built to answer. A vendor that charges per seat per month needs the tracking problem to be an everyone, every day, every hour problem. The larger the gap between perfect timesheets and your current timesheets, the stronger the case for the seat licence.

We see the same pattern in managed services, where the vendor math on ticket triage quietly assumes the volume being measured is the right volume to measure. And we see it in almost every automation ROI calculation we are handed: the denominator is chosen by the seller, and the denominator is where the argument lives.

Even the video results reflect it. Search the topic and you get tool comparisons and customer case studies. The one result that names the tension, Time tracking is not timesheets, puts it in the title and then argues it from inside a product (YouTube).

What changes when you ask the pricing question first

If the real job of time data is to price the next engagement, the requirements change shape, and they get smaller.

You do not need every hour, from every person, every day. You need enough data, of a trustworthy enough kind, to answer a short list of questions. Which engagement types consistently run over what we quote? Which clients generate revision rounds that were never in the scope? Which deliverables look identical in a proposal and cost twice as much in delivery? Where does the work go between the brief and the first draft?

Those questions have some useful properties. They are answered at the level of the engagement type and the client, not the individual. They can be sampled rather than collected exhaustively. And the answers are used to change prices, not to question people, which removes the reason staff shade the numbers in the first place.

This is the same move we recommend to manufacturers who lose money on work they quoted correctly. The cost data exists to make the next quote right. Once it becomes a tool for policing the last job, it stops being accurate enough to do either.

There is a harder conversation hiding here too. If your data, once it is honest, shows that a whole category of retainer is underpriced by half, the fix is not operational. It is a commercial decision about repricing, rescoping, or walking away from a client you like. No tracker makes that decision for you, and some agencies buy a tracker partly so they can keep not making it.

What we would not tell you to do

We would not tell you to stop tracking time. Agencies with hourly clients, regulated work, or government contracts need hourly records, and some clients contractually require them.

We would not tell you that the tracker you already own is the problem either. In most cases it is fine. It is being asked to serve two masters, a pricing model and a performance review, and it cannot serve both honestly.

And we would not tell you to buy a new one before you have decided what the data is for. That is the sequencing mistake the whole category is built on. A tool bought to recover leaked revenue on fixed-fee work will be measured against a revenue gain that cannot arrive, and in a year you will be having the Friday argument again, with a nicer interface. Time, it turns out, is not the only thing that gets logged and forgotten.

The pattern, stated plainly

The agencies that get this right tend to share one trait. They treat time data as an input to pricing, collected at the resolution pricing needs, and they keep it away from individual performance conversations. Their timesheets are less complete than their competitors' and considerably more accurate.

The ones that struggle measure utilisation per person, price per project by gut, and cannot connect the two. They know, to the percentage point, how busy each producer was last month. They cannot tell you which of their five service lines makes money.

Which of those you are is not something an article can tell you. It takes looking at a year of proposals next to a year of delivery records, and at how the numbers were entered, not only what they say.

If your time tracking keeps getting fixed and keeps coming back, it is worth checking whether the problem was ever the timesheet. Talk to us about how your agency prices its work, and we will tell you honestly whether a new tool belongs anywhere in the answer.