Nick QuirkYour agency tracked 412 hours last month. Your clients paid $58,000. Somewhere between those two numbers is the question nobody in your leadership meeting can answer: did the hours actually produce anything a client would pay for again?
Most agencies track time for billing and capacity. Very few track it in a way that connects to client value — and that gap is why profitable-looking retainers quietly churn. This article covers how to build the connection: a tagging system that survives contact with reality, the math that turns hours into value signals, and the reporting change that stops clients from counting your hours in the first place.
Why hours and value drift apart
Time tracking data drifts from client value for three predictable reasons.
The categories are built for accounting, not outcomes. "SEO — Client A" tells you nothing. A month of technical audit work and a month of hitting refresh on Search Console both land in the same bucket, and one of those bought the client a 12% organic revenue lift while the other bought a screenshot.
Time gets logged in retrospect. Friday-afternoon timesheet reconstruction produces round numbers that sum to 40 and describe nothing. We wrote more about this failure mode in why agency time tracking data lies, but the short version: if the data is invented, no analysis on top of it is real.
Nobody defined value on the client's terms. Agencies define value as effort delivered. Clients define it as change in their business or change in their confidence. Those are different, and only one of them renews the contract.
The four buckets every tracked hour falls into
Before you build anything, sort your existing time into four buckets. Pull last quarter's data and force every hour into one:
- Direct value work — produces an asset or change the client can point to: published content, implemented fixes, a migration plan, a strategy the client acted on.
- Enabling work — necessary but invisible: research, QA, competitor analysis, internal strategy sessions that inform direct work.
- Relationship work — calls, reporting, Slack, stakeholder management. Real value, frequently untracked, almost always underestimated.
- Leakage — rework, waiting on client approvals, scope handled "as a favor," tool fiddling, meetings about meetings.
A healthy retainer usually looks like 55–65% direct, 15–20% enabling, 10–15% relationship, and under 10% leakage. When we've run this exercise with agency teams, the surprise is rarely the direct number — it's that leakage runs 18–25% on their two or three most difficult accounts. Those are the same accounts the team complains about, which is not a coincidence. Unprofitable retainers are usually leakage problems wearing a pricing costume.
Step 1: Tag hours to deliverables, not activities
The single highest-leverage change: stop tracking to activities ("keyword research") and start tracking to named deliverables ("Q3 Cluster Map — Insurance Vertical").
Activities are infinite and uncomparable. Deliverables are countable, and countable things can be divided into other countable things. When a strategist logs 6.5 hours to a named deliverable, you learn what that deliverable actually costs. Do it twenty times and you have benchmarks:
- Technical audit, mid-size site: 14–18 hours
- Content brief with SERP analysis: 1.5–2.5 hours
- Monthly reporting package: 3–4 hours (if it's taking 7, the report is the product and it shouldn't be)
- Migration QA: 20–30 hours, and it will be 30
Keep the deliverable taxonomy under about 25 items across your whole service catalogue. Bigger than that and people pick the wrong tag out of fatigue. This is the piece most general PM tools get wrong — they're built for tasks, not for agency deliverables that repeat across dozens of clients. If you're evaluating options, our comparison hub breaks down how the major tools handle agency-specific structure versus generic task lists.
Step 2: Attach an outcome to every deliverable type
Each deliverable needs a defined outcome recorded when it ships — not a KPI dashboard, just one honest line. Three flavors work:
Measured outcomes
"Fixed 340 broken internal links; crawl budget on /products improved from 62% to 89% indexed in six weeks." Numbers with a before and after.
Decision outcomes
"Client killed the planned Spanish-language subdomain based on our analysis; saved an estimated $40K in dev spend." Strategy work often produces a decision, not a metric. Decisions are value. Record them.
Risk outcomes
"Caught canonical tags pointing to staging pre-launch." No metric moved, and that's the point. Prevented disasters need a place in the record or the work looks like it produced nothing.
This takes about 90 seconds per deliverable and it is the entire bridge between your timesheet and your renewal conversation. It also keeps you out of vanity-metric territory — see agency KPIs that actually matter for what's worth reporting versus what just looks busy.
Step 3: Run the value-per-hour math
Now the numbers connect. Here's a real-shaped example — a $6,500/month retainer with 42 hours logged:
- Technical fixes implemented: 11 hrs → 340 links repaired, indexation +27pts
- Content briefs (4): 9 hrs → 4 articles published by client's writer
- Strategy + roadmap revision: 5 hrs → client reallocated $18K of paid budget
- Reporting + calls: 8 hrs
- Rework on a brief that missed the audience: 4 hrs
- Waiting/coordination on dev access: 5 hrs
Effective rate is $155/hr. But 9 of those 42 hours (21%) produced nothing a client would name. Strip them and your direct-value rate is $197/hr — which tells you the work is priced fine and the process is where $1,400 of margin went. That's a different fix than raising prices, and you can only see it when hours are tagged to deliverables and deliverables to outcomes.
Run this across ten clients and patterns emerge fast: the client with three stakeholders costs 6 extra hours a month in coordination; content-heavy retainers deliver value at $180/hr while "SEO consulting" retainers deliver at $240/hr. That's a pricing roadmap, not a spreadsheet.
Step 4: Change what the client sees
Never send a client a timesheet. Send them a value summary with hours as supporting context:
"This month: 4 content briefs published, 340 technical errors resolved, and a paid-budget reallocation based on our channel analysis. 42 hours invested, 79% on direct build and fix work."
That paragraph does three jobs. It names outcomes, it proves effort, and it pre-empts the "what am I paying for" conversation without inviting line-item scrutiny. Agencies that report this way rarely get asked for hourly breakdowns — and when they do, the answer is already assembled.
If you're moving away from hourly logic entirely, this data is the prerequisite. You can't credibly quote value-based pricing without knowing your own delivery costs; the shift from hourly billing to value-based retainers depends on exactly the ben
