Marketing agency productivity is the reliability with which paid capacity becomes valuable client work at an acceptable delivery cost, quality, and pace. A full calendar or a high deliverable count cannot prove that a project was profitable. Rework, senior review, unpriced coordination, overtime, and unbilled revisions can all increase output while reducing margin.
To evaluate a $20,000 campaign, keep four values separate: what the client agreed to pay, what delivery cost, what was invoiced, and what was collected. These values answer different questions. This guide shows how to connect them to utilization, realization, capacity, scope control, and project-level decisions without treating a work-management platform as an accounting system.
No universal utilization or margin target fits every agency. Define the accounting basis, cost policy, time period, and project population before comparing teams, clients, or months.
Hours logged and deliverables shipped are inputs. Project economics show whether the work was priced, delivered, invoiced, and collected well enough to support the business.
Build a small metric set with explicit denominators
Use measures that reveal different parts of the operating picture. For every percentage, record its numerator, denominator, period, included projects or roles, accounting basis, and source system. Without those definitions, a precise-looking dashboard can produce misleading comparisons.
Utilization
Utilization = billable delivery hours divided by available delivery hours. Define available hours consistently. Start with scheduled or contracted hours according to a documented rule, then account for holidays, leave, training, management, internal work, and other non-delivery responsibilities that materially affect the role.
Utilization is a capacity and demand signal, not a measure of employee value. High utilization can conceal overtime or unsustainable workload. Low utilization can reflect weak demand, poor scheduling, or deliberate investment in sales and capability building.
Realization
Realization = invoiced value divided by the standard value of delivered billable work for the same period. Document the standard rate, discounts, write-off treatment, and invoice rules used in the calculation. Realization can fall when work is discounted, written off, left unbilled, delivered outside scope, or performed by a role mix that the original price did not support.
Project contribution margin
Project contribution margin = project revenue recognized under the agency’s accounting policy minus direct delivery costs, divided by that revenue. Direct delivery cost may include appropriately loaded employee labor, contractors, and direct external costs, but the agency must define the policy. This is project contribution margin, not net margin. Sales, leadership, finance, software, facilities, and other overhead sit elsewhere unless the accounting policy explicitly allocates them.
Revenue per FTE
Revenue per FTE = defined-period revenue divided by the agency’s FTE count for that period. Use it as a trend rather than a verdict. Media-heavy agencies should distinguish agency revenue from pass-through client spend when their accounting records support that distinction. Faster output improves economics only if the agency captures the saved time through pricing, reduced rework, or higher-value work.
Promethean Research reports a 13% average agency net margin for 2025. That is a report-level figure, not a project-margin target for every agency. Service mix, cost treatment, and accounting definitions limit direct comparison. AgencyAnalytics’ 2026 survey reports that 44% of agencies said clients expected faster turnaround, while 30% said clients perceived agency services as less expensive. These are survey findings, not operating targets.
Read utilization, realization, and project margin together:
- High utilization with low realization: the team is busy, but discounts, write-offs, unbilled work, or over-servicing may be eroding value.
- Low utilization with strong realization: the work may be economically attractive, but demand or scheduling may be insufficient.
- High utilization and realization with weak margin: pricing, role mix, labor costs, or delivery estimates may be wrong.
- Low utilization and low realization: investigate both demand and delivery economics before changing staffing.
Trace margin leakage through five project handoffs
A campaign estimated at 120 hours but completed with 155 logged hours is over estimate, but that variance is only a symptom. The useful question is where the economics changed. Was the revision limit unclear? Did added work lack approval? Was senior review omitted from the estimate? Did finance write off the overage?
For a retrospective, choose one recently closed engagement and compare its original fee, estimated and actual hours by phase, planned and actual role mix, included and delivered revisions, approved changes, direct external costs, delivered and invoiced work, and invoice and cash status. Investigate the largest evidence-backed variance first.
Ignition reports that 78% of respondents in its survey of 273 agency managers and executives rarely or only sometimes charge for out-of-scope work. This is a vendor survey result, not an audited estimate of industry losses. It does show why a scope decision and its commercial outcome should be recorded rather than left in messages.
Set a project-margin intervention gate before kickoff
Agree before work begins what counts as direct delivery cost, who owns the forecast, and what threshold requires review. A useful project record can include a stable engagement ID, fee and currency, estimated delivery cost, actual delivery cost, forecast remaining cost, target margin, scope-change status, and last source-sync time.
Use deterministic arithmetic for the forecast:
forecast margin = (project fee - actual delivery cost - forecast remaining delivery cost) / project feeFor example, these illustrative values produce a forecast margin of 31.5%. With a 35% internal target, the project routes to human review. The fields and workflow below are an editorial design, not a ClickUp-published template.
{
"project_id": "PRJ-1042",
"currency": "USD",
"project_fee": 20000,
"actual_delivery_cost": 7200,
"forecast_remaining_delivery_cost": 6500,
"target_margin_pct": 0.35,
"scope_change_status": "PENDING",
"forecast_margin_pct": 0.315,
"intervention_status": "REVIEW",
"source_last_synced_at": "2026-10-20T14:00:00Z"
}Validate that the fee is nonzero, the project ID exists, currency is consistent, and cost inputs follow the agreed direct-cost policy. Reconcile time and cost inputs before publishing a snapshot. A below-threshold forecast or pending scope change should route to the delivery lead and account owner. It should not automatically change a contract, invoice, write-off, or client commitment.
A margin forecast is a dated operating estimate, not a completed accounting result. Preserve the input snapshot, route the exception to a named owner, and leave invoices, payments, write-offs, and recognized revenue in the finance system of record.
Keep data at the correct grain. A time entry is an event. A scope change is an event. A project margin is a dated observation. A monthly report is an aggregate. Store source record IDs and observation times with snapshots. For concurrent imports, enforce a database unique key or use a transactional upsert. A lookup followed by insert is not sufficient because two workers can both find no existing record.
For example, an illustrative implementation might use source_system + source_event_id for source events, workspace_id + task_id + observed_at for task snapshots, and workspace_id + project_id + metric_date + aggregation_version for daily project aggregates. Do not use project_id alone as the unique key for events or snapshots. Citation records and AI runs need their own identity and review status.
Use ClickUp for delivery inputs, not as a substitute for finance
ClickUp documentation supports task fields such as assignees, dates, estimates, dependencies, relationships, and Custom Fields. Its API can return task data and task-level tracked time where entries exist. Custom Fields can store structured project context, but workspace-specific field and option IDs must be retrieved rather than invented.
ClickUp time tracking supports time entries with billable or non-billable status, descriptions, and tags. Its Billable Report Dashboard card can report billable time entries. Neither feature proves that time was invoiced, recognized as revenue, or collected. Labor cost rates, contractor costs, invoice status, write-offs, and payments require a finance source or controlled reconciled fields.
Use a stable engagement ID to associate tasks, time, scope events, and financial records. Require a project ID, date, duration, user, task, and billable classification where practical. Quarantine records missing required role, rate, currency, or period data. Reconcile task-level totals to the available reporting view before aggregating.
Relevant documentation includes ClickUp task fields and operations, task retrieval and tracked-time details, time tracking, and the Billable Report card. Plan and role availability should be checked before committing to a specific configuration.
Check capacity before promising a deadline
Nominal working hours are not available capacity. Before committing, account for leave, holidays, pitches, management, internal work, coordination, and current client commitments when they affect the required people and skills. Preserve contingency rather than planning every available hour.
A practical sequence is to enter estimated effort, assignee, start date, and due date; check configured capacity and assigned work; add represented internal commitments and known unavailable periods; then have the delivery or resource manager approve the date. ClickUp Workload view can visualize assigned work using measures such as time estimates and capacity limits. It cannot account for responsibilities that have not been represented in the workspace.
If the estimate, required skill, assignee, or availability is missing, return the request for planning. An empty calendar is not proof of capacity. Utilization is a role-specific planning signal, not a requirement to keep every person maximally booked.
See the official Workload documentation and Workload configuration guidance for the supported view and capacity settings.
Control scope changes with rules first and AI second
Set a baseline for approved scope and define the response when a request changes it: replace existing work, defer the addition to a later phase, or assess a change order and deadline. If the agency deliberately absorbs the work, record that commercial choice and its reason so the cost remains visible.
Use deterministic checks for missing project IDs or approval status, tasks added after the baseline, estimate overruns, duplicate source events, and unbilled time on closed work. A bounded AI role can summarize what appears to have changed and point the owner to relevant task or message evidence. The project or account owner decides whether to accept, remove, defer, price, or reschedule the work and then updates the system of record.
ClickUp’s Scope Creep Detector listing describes comparison with a designated baseline. It also says that distinguishing approved from unauthorized changes depends on the agency’s process. The listing does not establish contractual entitlement, change-order creation, invoicing, or duplicate prevention.
Make the first review evidence-led
Start with one project closed in the most recent quarter. Trace it from sold scope to cash, identify the largest explainable variance, and choose one intervention: revise an estimate, clarify a scope approval rule, add a capacity check, fix an invoicing handoff, or revisit pricing. Then measure the same variance on the next three to five comparable projects before changing the standard process.
- Original fee, currency, and estimated delivery cost.
- Estimated versus actual effort by phase and role.
- Planned versus actual role mix and direct external costs.
- Included versus delivered revisions.
- Approved changes versus additional work.
- Delivered work versus invoiced work, including write-offs.
- Invoice status versus collected cash.
Assign one owner to the largest variance, set a due date, and record the evidence behind the decision. Automate only after metric definitions, owners, source systems, validation rules, and exception routes are agreed. If you need help reviewing an existing task, time, and reporting setup, see a ClickUp audit. For implementation support after the operating rules are defined, explore ClickUp setup and automations.
The goal is not to make every person busier or to produce another dashboard. It is to make the path from contracted work to delivery cost, invoice, and cash visible enough that the right owner can intervene while a project can still be managed.