Agency client churn is usually high for one of three reasons, and when clients leave tells you which: inside 90 days it is the sale, at 3 to 9 months it is results the client cannot tie to revenue, after 9 months it is the relationship or budget. At 5% monthly churn, the average client stays just 20 months.
At a glance: diagnosing agency client churn
Agency client churn is diagnosed by tenure, not by exit-survey reasons, because clients almost always say “budget” whatever the real cause was.
- The metric: monthly logo churn = clients lost in the month ÷ clients at the start of the month. Average client lifetime in months = 1 ÷ monthly churn.
- The asset on this page: the Tenure Clock, a table mapping when clients leave to the likely cause, the test that separates it and the fix.
- The number to compare against: no agency churn benchmark we could verify at source exists, so compare against the Ramp Ratio below: client lifetime divided by months to first visible result.
- The cost: an agency of 24 clients at 8% monthly churn has to win about 23 new clients a year just to stay the same size.
- The quiet cause: the agency reports leads, the client counts sales, and nobody owns the gap between the two.
How it works
How to diagnose agency client churn
List every lost client
Pull the last 12 months of churned clients with start and end months. Add the fee each one paid.
Sort by tenure
Group exits into 0-3, 3-9 and 9+ months. The biggest cluster is where to look first.
Run the separating test
Compare proposals with month-3 reports, sample lead follow-up, or check contact changes. Each test rules a cause in or out.
Fix one window
Change the sale, the reporting or the relationship for that window. Re-measure churn as a rolling 12-month average.
MAKE MORE SALES.
Pay-Per-Result pricing — We scale sales HARD aligned to your interests, better than anyone else.
“Why is my agency’s client churn so high?” Start with the formula
Agency client churn is clients lost in a month divided by clients at the start of that month, tracked as a rolling 12-month average. An agency with 24 clients that loses one in a month has 4.2% churn that month; losing two makes it 8.3%. With samples this small, one month means little and six months means a lot.
Track two versions side by side. Logo churn counts clients; revenue churn counts the monthly fees they took with them. If revenue churn runs well above logo churn, you are losing your biggest accounts, which is a different problem from losing small ones. We could not find a public agency churn benchmark that we could verify at source on a stated denominator, so this page does not quote one. Your own trend, and the Ramp Ratio below, are the better yardsticks.
Want this done for you? We book qualified sales appointments on a Pay-Per-Result basis — you only pay for calls that actually land in your calendar.
What does high churn cost an agency? A worked calculation
High agency churn eats new-business capacity: at 8% monthly churn, a 24-client agency must win nearly two clients a month just to stand still. The inputs below are assumptions: 24 clients and a US$4,000 monthly retainer. Replace them with yours.
| Monthly churn | Average client lifetime | Clients lost per year | New clients needed per month to stay flat | First meetings needed per month (at 12.5 per client won) | Lifetime fees per client at US$4,000/month |
|---|---|---|---|---|---|
| 3% | 33.3 months | 8.6 | 0.72 | 9 | US$133,333 |
| 5% | 20 months | 14.4 | 1.2 | 15 | US$80,000 |
| 8% | 12.5 months | 23.0 | 1.92 | 24 | US$50,000 |
The 12.5 first meetings per client won uses the assumed 40% meeting-to-proposal and 20% proposal-win rates from our page on how marketing agencies get their own clients. Moving from 8% to 3% churn cuts the meetings needed to stay flat from 24 a month to 9, and the lifetime fees per client rise from US$50,000 to about US$133,333. Churn is the most expensive line on an agency P&L that never appears on it.
The Tenure Clock: when clients leave tells you why
The Tenure Clock sorts churned clients by how many months they stayed, because each cause of agency churn tends to surface at a different point in the relationship. No public dataset reports how often each cause occurs, so the rows are ordered by when they appear, which is also the order you can test them in. The windows are our proposed rules of thumb, not measured benchmarks.
| Left after | Likely cause | Test that separates it | Fix |
|---|---|---|---|
| 0–3 months | Sold the wrong expectation, or the wrong-fit client | Put the proposal next to the month-3 report. Did you promise an outcome the timeline could not deliver? Do early leavers share a lead source or package? | Write the 90-day definition of success into the proposal; stop selling the package that churns |
| 0–3 months | Slow or messy onboarding | Count days from signature to first live campaign for leavers vs stayers | A fixed onboarding checklist with a launch date agreed at signing |
| 3–9 months | Results the client cannot tie to revenue | Ask the client contact how much revenue your work produced last quarter. If they cannot answer, neither can their CFO | Report booked calls, pipeline and closed revenue, not only leads and cost per lead |
| 3–9 months | Leads go cold on the client’s side | Sample 20 leads and check when, and whether, the client first contacted them | Agree a follow-up standard, or fix follow-up as part of the service |
| 9+ months | Champion leaves, work moves in-house, or budget is cut | Exit reason “hired in-house” or a new contact in the last 90 days before exit | Relationships with at least two people at each client; a quarterly business review with the budget holder |
| Any time | Your own account manager turnover | Compare churn for accounts that changed account manager vs those that did not | Handover calls, and documented account history |
To run it, list every client lost in the last 12 months with the month they started and the month they left, then count how many fall in each window. A cluster in the first 90 days says the problem is in sales and onboarding, not delivery.
If we can’t make you money, we don’t deserve yours.
Pay-Per-Result pricing — performance-based alignment.
Is my churn structural? The Ramp Ratio
The Ramp Ratio divides average client lifetime by the months it takes a new client to see a result they value, and below 3 the churn is structural. A client who sees nothing for four months and stays an average of ten has a Ramp Ratio of 2.5: they spend 40% of their tenure paying for work that has not yet paid back. The threshold is our proposed decision rule, not a benchmark.
| Ramp Ratio | What it means | Fix first |
|---|---|---|
| Below 3 | Clients leave soon after the work starts paying back | Shorten time to first result: a quick-win deliverable in month 1, or a different channel mix |
| 3 to 5 | Normal friction; churn spread across causes | Run the Tenure Clock and fix the biggest window |
| 5 or more | Retention is not the constraint | Pricing, upsell and new-client acquisition |
SEO and content retainers usually have the longest ramp, paid media a shorter one; measure your own months to first result per service line rather than borrowing a number.
Why do agency clients leave when the leads look fine?
Agency clients leave when leads look fine because the agency reports leads, the client judges revenue, and slow follow-up on the client’s side sits between them. Sample 20 recent leads from one churn-risk account and check how long each waited for first contact and how many were ever reached. If most waited more than a day, the campaign is not the bottleneck.
You have three honest options. Coach the client’s team on a follow-up standard and report against it, which costs account-manager hours every month. Report booked calls instead of leads, which makes the gap visible but does not close it. Or take on the follow-up as part of the service, either by building a setter function in-house (hiring, scripts, calling and texting rules under the TCPA, and someone to manage it) or through a partner; our white-label lead generation for agencies page compares refer, resell and build, and AI appointment setting explains how booked calls are produced. How a client sees the same problem from their side is covered in “my agency isn’t delivering results”, and how billing on outcomes changes the conversation is in pay-per-result vs retainer.
Frequently asked questions
What is a good client churn rate for a marketing agency?
We could not verify a public agency churn benchmark at source, so we do not quote one. Use the arithmetic instead: at 3% monthly churn the average client stays about 33 months; at 5%, 20 months; at 8%, 12.5 months. Compare your lifetime with your time to first result using the Ramp Ratio.
How do I calculate my agency’s client churn rate?
Divide the clients lost in a month by the clients you had at the start of that month, then average the last 12 months. Track revenue churn alongside it, the monthly fees lost divided by the fees at the start of the month.
Is it cheaper to keep an agency client than to win a new one?
Usually, yes. A 2014 Harvard Business Review article reports that, depending on the study and the industry, acquiring a new customer is anywhere from five to 25 times more expensive than retaining an existing one. It also cites Bain & Company research by Frederick Reichheld that increasing customer retention rates by 5% increases profits by 25% to 95%. Neither figure is specific to agencies.
Why do agency clients say “budget” when they leave?
Budget is the polite exit reason, and it is true only when the client cannot see what the spend returns. Check the Tenure Clock: a “budget” exit at month five usually means results the client could not tie to revenue.
How many new clients does my agency need to replace churn?
Multiply your client count by your monthly churn. A 24-client agency at 5% monthly churn loses 1.2 clients a month, so it needs 1.2 new clients a month just to stay the same size, before any growth.
Pay-Per-Result appointments
See if we’re a fit
We book qualified sales appointments for you and you pay on results, not retainers. Our booking page asks a few quick questions so you find out in two minutes whether that model suits your business.
- 50,769+ appointments booked without cold calling.
- Pay-Per-Result pricing — you pay for booked, qualified calls.
- Pick your own time on our live calendar, no phone tag.