The Agency Profitability Stack: 5 Numbers to Track Weekly

02. Number 1: Utilization Rate

What it is: billable hours logged by the team ÷ total available hours × 100. Target range: 65-75% for most service agencies. Above 80% is a team strain warning. Below 60% is a revenue warning. What it predicts: next month's revenue, with approximately 60 days of lead time. A drop in utilization this week will show up in revenue in about two months — the earliest financial warning signal in the entire stack. When it goes wrong: below 60%, check whether the issue is pipeline (not enough client work), scoping (work is happening but not tracked against billable projects), or tracking (hours aren't being logged). Different causes, different responses. How to get the data: total billable hours from time tracker this week ÷ (team size × scheduled hours). Requires consistent daily time tracking.

03. Number 2: Effective Hourly Rate by Client

You have a client you've worked with for two years. Your standard blended rate is $150/hour. Last month they had 44 hours tracked and you invoiced $4,800 — which works out to $109/hour. The gap is $41 per hour, across 44 hours: $1,804 left on the table in a single month. Annualised, that's nearly $22,000 of work delivered below your market rate, and nobody flagged it because the invoice amount looked fine. Formula: total invoiced ÷ total hours tracked for that client in the same period. Run it monthly, per client. Target: within 10-15% of your standard blended rate. Clients consistently outside that range are underpriced, over-serviced, or both — and the effective rate number tells you which conversation to have first. When it goes wrong: the data rarely surprises you about which client. You probably already know. What it adds is the dollar figure that makes the conversation easier to have — and harder to keep avoiding. How to get the data: one table, one row per client, updated monthly. Invoice amount ÷ hours tracked. Four columns, fifteen minutes. The pattern becomes visible after two or three months.

04. Number 3: Gross Margin by Project

What it is: (project revenue − project labour cost) ÷ project revenue × 100. Labour cost = hours tracked × your blended labour cost per hour (salary + overhead, not the billing rate). Target range: 35-50% gross margin for most agency project work. Below 25% is a concern. Above 55% on a consistent basis usually means you're underinvesting in quality. What it tells you: whether each project is generating enough gross profit to cover overhead and produce net income. Revenue without margin is cashflow, not profit. When it goes wrong: below 25% — check whether the hours were over-estimated at billing, the scope expanded without billing adjustment, or the estimate was wrong. How to get the data: requires knowing your blended labour cost (not billing rate). Formula: total team salary cost ÷ total available hours = labour cost per hour. Apply to tracked hours per project.

05. Number 4: Retainer Hour Consumption Rate

What it is: hours consumed so far this period ÷ total hours in retainer × 100. Tracked mid-month to catch over-delivery before the invoice. Target range: at the midpoint of the month, you should be at approximately 50% of the retainer's hours. If you're at 75% with 15 business days remaining, you're on pace to over-deliver by 50%. What it tells you: whether you're on pace to deliver within scope for each retainer client. The value is in the timing — a mid-month check gives you two weeks to manage remaining scope before the month closes. When it goes wrong: the mid-month conversation — 'we're at 58 of 90 hours at the midpoint, here's what's remaining, do you want a change order or should we trim scope?' — is a 5-minute email that protects both sides. The end-of-month version, after the work is done, has no good options. You've either absorbed the overage silently (losing $X at your blended rate) or you're asking the client to pay for work they didn't approve in advance. Neither is a good look. The check on the 10th is worth it. How to get the data: current month's hours tracked per retainer client ÷ retainer hour allocation. Run this on the 10th-12th of each month.

06. Number 5: Pipeline Coverage Ratio

What it is: total value of active pipeline ÷ monthly revenue target. Active pipeline = proposals sent or in discussion, not signed. Weight by close probability if you track it. Target range: 3× monthly revenue target for a healthy pipeline. If your revenue target is $50,000/month and you have $150,000 in active proposals, you're at 3× coverage. If you have $60,000 in proposals, you're at 1.2× — dangerously thin. What it tells you: whether you're building next quarter's revenue today. Pipeline coverage is the furthest-forward-looking number in the stack — a thin pipeline today is a revenue problem in 60-90 days. When it goes wrong: below 2× coverage, business development urgency increases immediately. The response time between 'pipeline is thin' and 'revenue is thin' is 60-90 days. There's no fast fix — there's only an early warning. How to get the data: a simple CRM or deal list with value and stage for each active prospect. Total the values and divide by your monthly revenue target.

07. The 15-Minute Weekly Review

Reviewing all five numbers takes 15 minutes if the data is accessible. Here's the weekly sequence:

• Pull utilization rate for the week (5 min) — is it above 65%? Which team members are low?
• Check retainer consumption for all active retainers (3 min) — anyone on pace to over-deliver this month?
• Review any projects that closed this week for gross margin (3 min) — any outliers to understand?
• Update pipeline coverage if any deals moved (2 min) — is coverage above 3×?
• Flag any effective rate clients for the monthly review (2 min) — any consistently low clients to address?

08. What Good Numbers Look Like

A healthy agency profitability stack looks like this on a weekly snapshot:

• Utilization: 68-72%
• Effective hourly rate: within 10% of standard for all clients
• Gross margin: 38-45% on projects closed this week
• Retainer consumption: all clients at less than 55% at mid-month
• Pipeline coverage: 3.2× monthly target

09. Building the Dashboard

The dashboard doesn't need to be sophisticated. A spreadsheet with five rows, updated weekly, is sufficient. The columns: metric name, current week value, target range, status (green/yellow/red), action if red. What the dashboard needs to be is consistent. The same metrics, tracked the same way, every week. A metric that's measured inconsistently or defined differently week to week produces noise, not signal. The value comes from the trend over time, not from any single week's reading. If one of the five numbers requires more than ten minutes to pull together each week, that's a data infrastructure problem worth solving — because friction is what turns a weekly habit into a monthly one, and a monthly one into 'when I have time.'