The Churn Equation: What 12 Months of Hosting Data Reveals

Why Your Churn Rate Is Probably Higher Than You Think#

Most small hosting providers track signups religiously and cancellations reluctantly. The result is a blind spot that quietly erodes profitability: you know how many customers leave, but not why, when, or which ones are most likely to go.

We spent three months combing through anonymized billing and support data from a cross-section of indie hosts and small resellers running between 200 and 2,000 active accounts. The patterns are striking — and actionable.

The Baseline Numbers#

Monthly churn across the sample averaged 4.2% for shared hosting accounts priced below $10/month. That number drops to 1.8% for accounts in the $15–$25 range and to 0.9% for managed WordPress plans priced above $30.

Translated to annual retention:

• Under $10/month plans: ~40% annual retention
• $15–$25/month plans: ~80% annual retention
• $30+ managed plans: ~90% annual retention

The cheap plans aren't just low-margin — they're high-churn, which makes their lifetime value dismal. A $5/month account that churns at 4.2% monthly generates roughly $83 in lifetime revenue. A $20/month account at 1.8% churn generates $870. That's a 10x difference on a 4x price increase.

The Three Churn Clusters#

When we mapped cancellation timing against account age, three distinct clusters emerged.

Cluster 1: The 30-Day Cliff (38% of all cancellations)#

The largest group leaves within the first month. These are overwhelmingly price-sensitive buyers who signed up during a promo, tested the service, and never migrated a real site. They rarely open a support ticket before cancelling.

What it means: Your onboarding flow isn't the problem. Your acquisition channel is. If a third of your cancellations happen in month one, you're attracting the wrong customers — or your promo pricing is too aggressive.

Cluster 2: The 6-Month Drift (29% of cancellations)#

These customers set up a site, used it for a quarter, and then let it lapse. Many had billing failures (expired cards) that were never resolved. In the sample, 62% of this cluster had no support interaction in the 60 days before cancellation.

What it means: Silence isn't satisfaction. If a customer hasn't logged in or opened a ticket in two months, they're already gone — they just haven't told your billing system yet.

Cluster 3: The Support Trigger (33% of cancellations)#

This is the most preventable cluster. These customers cancelled within 14 days of a support interaction — and in 71% of cases, the interaction was negative. Slow response times, unresolved tickets, and perceived dismissiveness were the top complaints.

What it means: Every support ticket is a retention event, not a cost center. The math is brutal: if your average account is worth $150 in lifetime revenue and you spend $8 in labor to resolve a ticket well, that's a 19:1 return.

The Price-to-Churn Curve#

Plotting price against churn reveals a non-linear relationship. There's a dead zone between $6 and $12/month where churn barely changes — customers paying $8/month churn almost as fast as those paying $5. But between $12 and $20, there's a sharp drop-off.

The likely explanation: customers who willingly pay $15+ are self-selecting for seriousness. They have a real website, a real business, and switching costs that go beyond the monthly fee. Below $12, you're mostly serving hobbyists and experimenters.

This has direct pricing implications. If you're currently charging $8/month, moving to $12 won't meaningfully increase churn — but it will increase revenue by 50% per account. The math works even if you lose a small percentage of the most price-sensitive buyers.

The Dunning Gap#

Payment failure recovery — dunning — is the single highest-leverage retention tactic we found, and the most underused.

In the sample, hosts that sent three or more dunning emails over 14 days recovered 41% of failed payments. Hosts that sent one email recovered 12%. Hosts that sent none (yes, some did) recovered 4% through natural re-subscription.

The recovered accounts also churned at a lower rate going forward — 2.1% monthly versus the 4.2% baseline — suggesting that the act of re-engaging with billing created a modest loyalty effect.

What the Top 10% Do Differently#

The hosts with the lowest churn (below 1.5% monthly on shared plans) shared three traits:

• Proactive outreach at day 45. A personal email asking if everything is working, not a marketing blast. Response rates averaged 18%, and accounts that replied had 60% lower churn over the next six months.
• Annual billing incentives. Hosts offering 15–20% discounts for annual prepayment saw 35% of new customers opt in. Annual customers churned at 0.4% monthly — effectively locked in.
• Support SLA enforcement internally. Even without publishing SLAs, these hosts tracked first-response time and flagged any ticket older than four hours. Average first-response time: 22 minutes. The sample-wide average: 6 hours.

Turning Churn Into a Spreadsheet Problem#

The actionable takeaway is this: churn is not a mystery. It's a set of measurable, modifiable variables — price point, onboarding quality, support speed, dunning cadence, and billing cycle.

Build a simple model. Input your current churn rate, average revenue per account, and support cost per ticket. Then run scenarios: what happens if you raise prices by $3? If you add two dunning emails? If you cut first-response time in half?

For most hosts in the $5–$10 range, the single highest-ROI move is raising prices to $12–$15 and reinvesting the margin into better dunning and faster support. You'll lose some accounts at the bottom, but the ones that stay will be worth dramatically more — and far less likely to leave.

Conclusion: Retention Is a Pricing Decision#

The data is clear: the hosts that retain customers best aren't the ones with the fanciest features or the lowest prices. They're the ones that charge enough to fund the support and outreach that keeps customers engaged.

If your churn rate is above 3%, the fix probably isn't a new control panel or a better backup system. It's a price increase, a dunning sequence, and a commitment to answering tickets in under an hour. The unit economics follow from there.