Most agency owners have a reporting problem they do not realise they have. It is not that they lack data — it is that they are measuring the wrong things. Website traffic, social followers, email open rates, number of tasks completed: these numbers feel productive to report on, but none of them tells you whether your agency is actually healthy. This article is about the seven metrics that do.
The Vanity Metric Trap
Vanity metrics are seductive because they tend to go up. Traffic grows. Followers accumulate. Tasks get ticked off. Reporting on them feels like progress because the numbers improve even when the underlying business does not. The problem is that none of them has a direct, mechanistic relationship with agency profitability or longevity.
The agencies that run into trouble — slow cash-flow crises, sudden churn, margin erosion that only becomes visible at year-end — are almost always tracking activity rather than outcomes. They know they logged 2,400 billable hours last month but not what margin those hours generated. They know a client renewed but not whether that client is quietly dissatisfied and will churn at the next opportunity. They know their pipeline has eight open opportunities but not the realistic weighted value of those opportunities.
Good agency reporting is not about generating more data. It is about identifying the smallest set of numbers that, taken together, give you an accurate and timely picture of business health. Here are the seven that belong in that set.
1. Billable Utilisation Rate
For delivery staff. Below 65% is a margin problem. Above 88% is a burnout risk that will eventually become a quality problem.
Billable utilisation is the percentage of your team’s available working hours spent on work that gets invoiced to clients. It is the single most diagnostic number in a service business. If your pricing model requires 75% utilisation to hit margin targets and your team is running at 61%, that 14-point gap explains most margin problems without any further analysis. No spreadsheet required.
Calculate it correctly: billable hours ÷ contracted available hours × 100. Contracted available hours means total working hours minus confirmed leave — not an arbitrary 40-hour week applied uniformly. A part-time designer working 3 days per week has 24 available hours, not 40. Getting this denominator wrong produces a flattering but misleading figure.
Track utilisation by individual, not just team aggregate. A 76% team average can conceal a senior developer at 93% — overloaded, making mistakes, close to burnout — sitting next to a junior at 51% who is insufficiently challenged or poorly allocated. The individual breakdowns are where the actionable intelligence lives. Marque’s utilisation reporting surfaces both views directly from logged time entries, so you can see the pattern without manually compiling anything.
2. Gross Margin Per Client
Any retainer client delivering under 40% gross margin after direct delivery costs deserves a pricing conversation. Under 25% is almost certainly loss-making once you factor in account management time.
Revenue per client is easy to see. Gross margin per client requires more work, which is why most agencies do not look at it — and why they are often surprised when a long-standing, apparently healthy client turns out to be unprofitable. The calculation is straightforward: monthly recurring revenue minus the direct cost of delivering that client’s work (staff time at your internal blended rate, plus any freelancer or tool costs billed specifically to that account).
For a 10-person agency with a blended delivery cost of £55 per hour, a £2,500/month retainer that consumes 38 hours of delivery time is costing £2,090 to deliver — a 16% gross margin before you account for account management, reporting, or client calls. That same £2,500 retainer consuming 22 hours costs £1,210, generating a 52% margin. Both clients look identical on a revenue report. They are profoundly different businesses.
Run this analysis quarterly and rank clients by gross margin. The bottom quintile warrants scrutiny. Some will be loss-makers due to scope creep you have not billed for. Others will have been under-priced at the outset. A few will be structurally difficult relationships that consume disproportionate account management time. Each requires a different response, but none can be addressed until the number is visible. See our guide to tracking profitability per client for the full methodology.
3. Monthly Recurring Revenue and Its Trend
Absolute MRR matters less than its direction. A flat MRR at £45k/month with rising delivery costs is quietly shrinking in real terms.
Monthly recurring revenue — the predictable, contracted income arriving each month regardless of new business won — is the foundation of agency financial stability. It is what lets you hire with confidence, plan capacity accurately, and sleep on a Sunday night without worrying about whether this month’s salaries are covered. Project revenue is fine. Retainer MRR is what scales.
The metric to track is not just MRR but its month-on-month change, broken down into three components: new MRR added (new retainers signed), expansion MRR (existing clients upgrading or expanding scope), and churned MRR (retainers lost or reduced). An agency growing at £1,500 of new MRR per month while churning £2,200 is shrinking, even though its new business team appears active. Tracking these three components separately tells you where the problem actually is.
Target a net MRR growth rate of 5–8% per month if you are in early growth phase, settling to 2–3% per month as you reach the size you want to operate at. If you are using retainers as your primary revenue model, this number is your most important financial metric by some distance.
4. Client Health Scores
A client whose health score drops 20+ points over a four-week period is displaying classic pre-churn behaviour. Intervention at this point saves the relationship significantly more often than waiting for renewal time.
Churn is the most expensive event in an agency’s operating life. The visible cost is the lost MRR. The hidden costs — account management time on the exit conversation, the morale impact on the delivery team, the gap in capacity while you replace the revenue, the disruption to forward planning — are often two to three times larger than the direct revenue loss. And most churn is not sudden. It telegraphs itself weeks in advance through a pattern of signals that nobody is watching for.
Client health scores aggregate those signals into a single number. The inputs vary by agency, but a robust model typically includes: project delivery against deadline (are projects consistently late?), invoice payment behaviour (getting slower to pay is a reliable churn predictor), support ticket volume and resolution time, recency of meaningful client contact, and direct satisfaction signals if you run periodic check-ins. An account scoring 82 last month and 61 this month has changed in some meaningful way. That change is worth a proactive conversation now rather than a reactive one in six weeks when the client decides not to renew.
Marque CRM’s client health score system calculates this automatically from activity data already in the platform — no separate survey tool, no manual scoring. Accounts that drop below a configurable threshold trigger a notification to the account manager. That is the difference between a system that measures health and one that actually improves retention.
5. Average Project Margin vs. Budget
If you are consistently delivering projects in 30% more hours than you quoted, your pricing model has a systematic error that compounds over every project you win.
Every project you run is, in effect, a small business experiment. You quoted a price based on an estimated effort. You delivered the project at an actual effort. The difference between those two numbers — quoted hours versus actual hours — is one of the most instructive data points in agency management, and most agencies either do not track it or only look at it when something has gone badly wrong.
If your average project comes in 15–20% over the quoted hours, you have a scoping problem. Your estimators are consistently optimistic, probably because they are quoting based on best-case assumptions rather than historical reality. If particular project types consistently overrun — website builds always run 25% over while SEO campaigns run on budget — you have a category-specific pricing issue that can be corrected by adjusting your rate card or your discovery process for those project types.
The remedy is to build a feedback loop. After every project closes, compare quoted to actual hours. After five or ten projects of the same type, you will have a reliable correction factor. Apply it at quoting stage. This is not sophisticated project accounting — it is basic operational learning that most agencies skip because they close a project and immediately move on to the next one without looking back. Our guide on tracking project budget profitability covers this process in detail.
6. Weighted Pipeline Value
An unweighted pipeline of £80k in open opportunities is a story. A weighted pipeline of £24k tells you what you can actually plan around.
Most agency pipelines are a list of things the team hopes will happen. A prospect had a good first call — that goes in the pipeline at the full value of the potential contract. Four months later, it is still there, still at full value, even though the prospect has gone quiet and the opportunity is probably dead. The result is a pipeline that systematically overstates future revenue and makes capacity and hiring decisions look safer than they are.
Weighted pipeline value multiplies each opportunity by a realistic close probability based on its current stage. An initial enquiry might be weighted at 10%. A proposal submitted and under review might be weighted at 35%. A verbal agreement with a contract outstanding might be at 80%. Apply those weights consistently and your pipeline number becomes a genuine planning tool — you can look at £31k of weighted pipeline for the next 90 days and make a real decision about whether you need to actively generate more business or whether the team is covered.
Review your pipeline weekly, but avoid the common mistake of adjusting probabilities upward based on optimism rather than evidence. A deal that was 35% three weeks ago and has had no meaningful movement since is not now 50% just because time has passed. Stage-based probabilities should be driven by what the prospect has done, not what you hope they will do.
7. Average Support Ticket Resolution Time
For agencies on retainer with an SLA commitment, first response within business hours and resolution within 24–48 hours is the baseline clients expect. Slipping on this is one of the fastest ways to erode a relationship.
Support quality is not a soft metric — it is a direct predictor of retention. When a client raises a support ticket and it sits unacknowledged for 18 hours, or gets passed between team members without resolution, or receives a superficial reply that does not actually address the issue, that client is logging data. Their mental model of your agency is being updated with each interaction. After three or four poor support experiences, the renewal conversation becomes very difficult.
Tracking average first response time and average resolution time across all client tickets gives you an objective view of support quality that is independent of individual client perceptions. You want both numbers moving down over time, and you want them segmented by client — because an SLA commitment in a contract is not meaningful if it applies on average but some clients are consistently underserved.
Agencies that handle support through email chains or Slack DMs cannot track these metrics reliably, because the data does not exist in a structured form. A proper support ticketing system — ideally one integrated with your CRM so you can see ticket history alongside the full client relationship — makes this reportable by default. Marque’s built-in support ticketing includes SLA tracking, first response measurement, and client-level reporting without requiring a separate tool like Zendesk or Freshdesk.
Building a Reporting Cadence That Actually Gets Used
The seven metrics above are only useful if you look at them consistently. Sporadic reporting — pulling numbers when something feels wrong — misses the point. You need a cadence that makes anomalies visible before they become problems.
A practical structure for a small agency looks like this: a five-minute daily review of support ticket queue and anything flagged urgent; a 20-minute weekly review of utilisation by person, weighted pipeline, and any client health scores that have moved significantly; a 45-minute monthly review of MRR trend, gross margin per client, and project margin vs. budget across all closed projects in the period. This is not a huge time commitment. What it does is replace the end-of-quarter surprise — “how did we end up here?” — with a steady flow of small course corrections made at a point when they are still inexpensive to execute.
The prerequisite is that your operational data is in one place. If your time tracking lives in Toggl, your invoices in Xero, your projects in ClickUp, and your client communication in email threads, you cannot build a coherent view of any of these metrics without significant manual effort. That manual effort is what kills reporting cadences. The person responsible runs the numbers once, finds it takes two hours, and never does it again. The solution is not better discipline — it is better tooling.
A platform like Marque CRM connects time tracking, invoicing, project delivery, support tickets, and client communication in a single system. The metrics described in this article are not reports you have to build — they are views of data that already exists because your team is working inside the same platform. That is the structural change that makes weekly agency reporting a five-minute task rather than a half-day project. See our pricing page to understand which plan tier includes which reporting features.
Where to Start
If you are currently tracking none of these metrics, do not try to implement all seven at once. Start with utilisation rate and MRR trend. These two numbers, tracked weekly, will surface more actionable information in 30 days than most agencies get from months of more elaborate reporting.
Once those two are habitual, add gross margin per client as a monthly calculation. Then introduce client health scores if your platform supports them. Build toward the full set over a quarter, not a sprint. The goal is not a comprehensive dashboard for its own sake — it is a reporting habit that permanently improves the quality of the decisions you make about your agency.
The agencies that grow deliberately, hit their margin targets, and retain clients for years are not doing something magical. They are looking at the right numbers, at the right frequency, and making small adjustments before small problems become large ones. That discipline is replicable. It just requires knowing which numbers to look at.