
Aug 11, 2026
Written by Gregory Shein, CEO & Founder
The Agency Operations Dashboard: 9 Metrics to Review Weekly (and Who Owns Each)
Agency dashboards fail in two directions. Most shops have no dashboard at all — the owner's "dashboard" is a bank balance and a feeling. The rest over-correct into a 40-widget wall of charts that everyone admires at the quarterly review and nobody reads on a Tuesday.
A working operations dashboard is neither. It's a small set of numbers — nine, in the version below — reviewed weekly, where each number has a named owner, a healthy range, and a defined action when it leaves that range. A metric without an owner and a threshold isn't a metric; it's decoration.
This is the full list, who looks at what and when, a sample layout you can copy, and a worked read-through showing how the dashboard catches a revenue hole six weeks before it reaches the bank account.
The 9 numbers
Money coming (1–3)
1. Weighted pipeline coverage — sum of (deal value × stage probability) ÷ next-90-day revenue target. Healthy: ≥ 3×. Below 2.5× means a revenue hole is already forming; the lag from "coverage dips" to "invoices dip" is your average sales-cycle plus delivery lead time — usually 6–10 weeks. This is the single most ignored number in small agencies, because it predicts pain instead of describing it.
2. New-business activity — outreach touches, proposals sent, first calls booked this week. This is the input behind pipeline coverage: coverage tells you there's a problem, activity tells you why. If you track lead sources properly (the outreach system exists for exactly this), you also get win rate per channel, which turns "do more marketing" into "do more of channel X."
3. Revenue month-to-date vs target — invoiced + scheduled-to-invoice this month against plan. Simple, but it forces the invoicing to actually happen (metric 9's best friend).
Work happening (4–6)
4. Billable utilization — billable hours ÷ available hours, per person and team. Healthy for delivery roles: 65–80%. Below 60%: you're overstaffed or underselling. Above 85% for weeks: you're borrowing from next quarter (burnout, no improvement time). Benchmark yourself with the billable utilization calculator.
5. Capacity available, next 4 weeks — bookable hours minus committed hours, by role. This is the forward twin of utilization: utilization looks backward at last week; capacity decides whether sales may keep selling and when the next project can start. The agency capacity calculator gives you the baseline math.
6. Project health (RAG count) — how many live projects are green / yellow / red on burn ratio (hours consumed vs work complete). The dashboard wants the count; the per-project drill-down is its own weekly ritual, covered in project profitability tracking.
Money banked (7–9)
7. Delivery margin, month-to-date — (recognized revenue − time cost at loaded rates) ÷ revenue. Healthy: 50–60% delivery margin, which lands ~15–25% net after overhead. Falling margin with stable revenue = leakage: unbilled hours, scope drift, or a client that needs the by-client P&L treatment.
8. Effective hourly rate — revenue ÷ all client hours (billable or not), blended. The lie-detector metric: rate cards say $110; effective rate says what you actually earned per hour of attention. A widening gap between the two is scope creep wearing a trench coat.
9. AR outstanding + % overdue — total unpaid invoices, share past due, and oldest invoice age. Healthy: overdue < 15% of AR, nothing over 45 days. Revenue you've earned but not collected is the most expensive loan you'll ever make.
Who looks at what, and when
| Ritual | When | Who | Metrics | Output |
|---|---|---|---|---|
| Delivery stand-up | Mon, 15 min | PM + leads | 4, 5, 6 | Reassignments; red projects get an owner + action |
| Pipeline review | Tue, 20 min | Owner + sales | 1, 2 | Channel focus for the week; stalled-deal pushes |
| Money check | Fri, 15 min | Owner + ops/bookkeeper | 3, 7, 8, 9 | Invoices out; collection calls assigned |
| Full sweep | Monthly, 60 min | Everyone above | All 9 + trends | Rate, staffing, and client-mix decisions |
Note what's absent: no daily anything (weekly numbers wobble daily; you'll chase noise), and no metric reviewed by "the team" — plural owners means no owner.
Sample dashboard layout (copy this)
One screen, three bands, red/amber/green against the thresholds:
| Band | Widget | This week | Trend (8w) | Owner | Threshold |
|---|---|---|---|---|---|
| SELL | Pipeline coverage | 3.4× | ▁▂▃▄▄▅▅▅ | Owner | red < 2.5× |
| Proposals sent / calls booked | 5 / 3 | steady | Sales lead | red < 3/wk | |
| Revenue MTD vs target | 84% | on pace | Owner | red < 80% at wk 3 | |
| DELIVER | Utilization (team) | 74% | 71→74% | PM | amber > 80%, red < 60% |
| Capacity next 4 wks | 210 h | shrinking | PM | red < 1 project slot | |
| Projects R/A/G | 1 / 2 / 6 | stable | PM | red ≥ 2 red | |
| COLLECT | Delivery margin MTD | 54% | 57→54% | Owner | red < 45% |
| Effective rate | $86 | $92→$86 | Owner | red < 0.75× rate card | |
| AR / % overdue | $48k / 11% | improving | Ops | red > 20% overdue |
Worked read-through: the quiet week that wasn't
Here's a real-shaped snapshot from a 12-person agency, and what a 30-minute weekly review extracted from it:
- Utilization 82%, capacity next 4 weeks down to 90 hours, all projects green. Delivery feels great — high-fives all around.
- Pipeline coverage 1.9× (was 3.1× six weeks ago); proposals sent last week: 1.
- Margin 56%, AR clean. Money looks fine.
The naive read: "busy and profitable, no action needed." The dashboard read: this agency is six weeks from a trough. Everyone is heads-down delivering (82% utilization is why only one proposal went out — selling stopped), and at 1.9× coverage, when current projects roll off there's nothing behind them. The math: next-quarter target $180k, weighted pipeline $342k... at their 30% close rate on weighted value, that's ~$100k of likely revenue — a $80k hole, visible today, invisible in the bank account until October.
The actions, decided that Tuesday: owner blocks 6 hours/week for outbound (utilization target lowered to 75% — deliberately), two warm past-client threads reopened, one project's start pushed a week to free a senior for two proposals. Cost of the fix in week one: trivial. Cost of discovering the same hole in October: two months of payroll on reserves.
That's the whole value proposition of the dashboard: it converts "we're slammed" from a mood into a leading indicator.
Build it where the data already lives
Every metric above is simple division. The reason agencies don't have this dashboard isn't math — it's that the numerators and denominators live in four tools: deals in a CRM, hours in a tracker, budgets in spreadsheets, invoices in an accounting app. Assembling nine numbers weekly from four exports takes ~2 hours, so it survives three weeks and dies.
The durable version is structural: when deals, projects, tracked time, and invoices share one data model, all nine numbers are queries against data that's already joined — utilization, burn, margin, coverage, and AR update themselves, and the weekly ritual is 30 minutes of deciding instead of 2 hours of compiling. That's how Corcava's reporting works across the agency lead-to-invoice workflow: the dashboard is a byproduct of running the work, not a second job.
Stand up your nine numbers this week — start a free 14-day Corcava trial (no credit card), and begin with the two that predict the future: check your billable utilization and your pipeline coverage today.