Nick QuirkMost agency time tracking rollouts fail in the same way. The owner announces it on a Monday, the team logs diligently for three weeks, entries get vaguer ("client work — 6h"), and by month three someone is backfilling a whole week on Friday afternoon. The data is now worse than useless, because you're making pricing decisions on fiction.
The failure isn't the tool. It's that nobody told the team what the data was for, and the team correctly assumed the worst: that it was for judging them.
Why agency time tracking feels like surveillance (and when it actually is)
Time tracking is demoralizing when the numbers flow upward and nothing flows back down. People log 37 hours, hear nothing, then get asked in a 1:1 why Tuesday looks light. That's surveillance, and the team is right to resent it.
It stops feeling that way when the data visibly changes decisions that affect the people logging it: a client gets re-scoped, a bad-fit account gets fired, headcount gets approved, a deadline moves. The logging effort buys them something.
Here's the honest test. Ask yourself: in the last six months, what decision did I make because of timesheet data? If the only answer is "I checked whether people were working enough," you have a trust problem, not a tooling problem, and no software will fix it.
The three legitimate reasons to track time
- Pricing. You cannot price a retainer you can't cost. If you charge $5,000/month for local SEO and it eats 48 hours at a $95 blended cost, you're making $440. You need to know that.
- Scope defense. "This quarter you asked for 14 extra landing pages and three ad-hoc reports — that's 22 hours above the retainer" is a conversation you can only have with records.
- Capacity. Knowing that technical audits reliably take your senior 11-14 hours lets you say no to the fifth one this month before it blows up. This is the same data that drives agency capacity planning.
Notice what's not on the list: measuring individual productivity. If you need timesheets to tell you whether someone is performing, your management is broken upstream.
Design the system around the smallest useful unit
The single biggest driver of timesheet resentment is granularity. Agencies that ask for six-minute increments against 40 task codes get garbage data, because nobody can reconstruct their day that precisely and everyone rounds to the nearest plausible story.
Track client × phase, nothing finer
For an SEO agency, 6-10 categories covers everything you'll ever act on:
- Technical (audits, implementation support, crawl work)
- Content (briefs, editing, publishing)
- Links / digital PR
- Local / listings
- Reporting & analysis
- Client comms (calls, email, Slack)
- Internal (admin, training, sales support)
Fifteen-minute increments. If a day looks like "Acme — Technical — 3.5h, Acme — Client comms — 0.75h, Internal — 1h," you have everything you need. Mapping those categories to the delivery phases you already use keeps it almost automatic; if your workflow follows a structured model like the one in project management for SEO agencies, the time categories should just mirror the phases.
Make it take under two minutes a day
Measure the cost of compliance honestly. If logging takes eight minutes a day, that's 32 hours a year per person — roughly $3,000 of cost for a team of six. The system has to pay for that.
Practical fixes: timers started from the task the person is already working in, a weekly grid view where last week's pattern can be duplicated and adjusted, and no mandatory notes field. Required free-text notes are where compliance goes to die. This is one of the things worth checking when you compare agency PM tools — whether tracking is attached to the work itself or sits in a separate app the team has to remember to open.
What to never do with the data
Three commitments, stated out loud at rollout and honored without exception:
Never use timesheets in a performance review. Not as evidence, not as context, not "I noticed." The moment hours appear in a review, every number after that is negotiated rather than recorded.
Never compare individuals publicly. Compare accounts. "Acme costs us 54 hours a month and Beta costs 21 for the same fee" is a business fact. "Priya logged 31 billable hours and Tom logged 38" is a weapon.
Never ask about a single day. Noise at the day level is enormous. If something concerns you, look at a four-week rolling average, and even then ask about workload, not about hours.
One agency owner I know prints the rule on the timesheet page itself: "This data is used for pricing and scoping. It is never used in reviews." That costs nothing and changes the psychology of every entry.
The rollout: four weeks, not one announcement
Week 1 — You track, alone
Before you ask anyone else, track your own time for a week at the granularity you're about to mandate. You'll discover which categories are ambiguous, how long it actually takes, and at least one assumption about your own week that's wrong. You'll also have standing when someone complains.
Week 2 — Explain with a real number
Don't open with "we're implementing time tracking." Open with the problem: "I think the Beta retainer is underwater and I can't prove it. Here's what I want to find out, here's exactly how I'll use it, here's what I will never use it for."
Then set the expectation that the first month's data will be messy and that's fine. You're building a baseline, not auditing.
Week 3-4 — Log, don't police
No chasing. If entries are missing, note it and move on. The goal of month one is habit, not completeness.
Week 5 — Give the data back
This is the step everyone skips and it's the one that determines whether the system survives. Share, with the whole team:
- Hours per account vs. fee, as a simple list
- Which accounts are profitable and which aren't
- What you're going to do about the unprofitable ones
Then do the thing. Re-scope one client, or raise one price, or fire one account. The team now has direct evidence that their logging produced a decision that protects their workload. That single moment is worth more than any amount of policy.
Reading the numbers without drawing the wrong conclusion
Set a utilization target, not a maximum
A realistic billable target for delivery staff is 60-75% of a 40-hour week — 24 to 30 hours. The remaining time is genuinely productive: internal reviews, learning, documentation, covering for colleagues. Agencies that push for 85%+ get it for one quarter and then get burnout and turnover, which costs far more than the margin gained.
Publish the target. People who don't know what "good" looks like will assume "more."
Use medians and ranges, not averages
One migration-heavy month will drag an average sideways. If your technical SEO audits come in at a median of 12 hours with a range of 9-16, that's a usable estimate. An average of 13.4 across four audits, one of which was a 45-page enterprise site, is not.
Investigate variance before you blame it
When an account runs 40% over estimate, the cause is almost always one of four things: unclear scope at sale, a client who sends work through three channels, a deliverable nobody templated, or genuine under-pricing. Three of those four are your problem, not the team's. Ask "what made that slow?" and you'll usually get a specific, fixable answer within a minute.
Turning hours into client conversations
The highest-return use of time data isn't internal — it's the quarterly scope conversation. "Your retainer assumes 30 hours. We've averaged 41 for two quarters, mostly on ad-hoc requests. Two options: we raise the retainer to match reality, or we agree that ad-hoc requests go into a separate bucket." Clients respect this. It reads as competence, not complaint.
Pair that with the kind of structured evidence covered in SEO client reporting and the conversation shifts from price defense to value demonstration. Tools built for agency workflows — PeakKR among them — tie hours to phases and deliverables specifically so these conversations have receipts behind them.
Signs your system has gone wrong
- Entries cluster on Friday afternoon — people are reconstructing, not recording
- Every account lands suspiciously near the budgeted hours — the team is logging to target
- Internal and admin time has quietly vanished — it's being hidden in client codes
- Someone asks "does this count as billable?" more than once a month — your categories are unclear
- You've collected six months of data and changed nothing — the team has noticed, even if they haven't said so
The rollout checklist
- Write down the three decisions you intend to make with
