One of the most interesting indications of churn is the one most teams read backwards.
If there are no support cases, the customer is at risk of churn.
The shape
Support ticket volume does not predict churn in a straight line. It is U-shaped, and both ends are bad.
| Tickets in the last year | Churn risk | What it usually means |
|---|---|---|
| Zero | Elevated | Not using the product, or not deeply |
| Three to five, tier 2 or 3 | Lowest | Engaged, supported, usually happy |
| Seven or more | Elevated | Persistent unresolved friction |
Why zero is a warning
The intuitive reading of silence is that the product is working. Occasionally that is true. More often it means nobody is using it hard enough to find the edges.
Real usage generates questions. A team that has integrated a product into how it works will hit a permissions question, an edge case in a report, a request to do something the documentation does not cover. The absence of any of that over a year is not a testament to product quality, it is an absence of depth.
Silence also removes the relationship. Support interactions are contact points where value gets demonstrated and where a human at the vendor learns what the customer is trying to do. A customer with zero tickets has had no such contact, so at renewal there is no accumulated evidence of value and nobody on either side who can speak to it.
This mirrors the pattern in deal slippage, where the earliest warning is the absence of a signal rather than the presence of a bad one. Systems built to inspect what is present are structurally poor at noticing what is missing.
Why seven or more is also a warning
The high end is the reading most teams already have, and it is correct for a different reason. A customer generating a high ticket volume is hitting friction repeatedly, and each unresolved ticket is a small piece of evidence that the product is costing them time.
The severity distribution matters here. Three to five tier 2 or tier 3 tickets look like a customer working through normal questions. A comparable count of severe tickets is a different situation entirely, and volume alone will not distinguish them.
Building the signal
The data usually already exists, in a system the revenue team does not look at.
1. Join support ticket counts to the account record over a trailing twelve months. This is often the entire project, and it is frequently the step that has never been done because the systems belong to different teams. 2. Band the counts rather than using a raw number, so that the U-shape is expressed rather than averaged into a meaningless linear coefficient. 3. Split by severity, so five routine questions are not scored the same as five outages. 4. Combine with usage data where you have it. Zero tickets plus falling usage is a much stronger signal than zero tickets alone, because it separates the disengaged customer from the genuinely self-sufficient one.
What to do at each end
The two ends need opposite interventions, which is the practical reason the U-shape matters rather than being a curiosity.
Zero-ticket accounts need proactive outreach, and specifically the kind that surfaces whether anyone is actually using the product. A check-in that asks whether everything is fine will get a yes. A conversation about what they have built with it will not. High-ticket accounts need resolution and escalation, and an owner who can see the pattern across tickets rather than closing them individually. Seven separate tickets closed successfully can still add up to a customer who has decided the product is difficult. Middle-band accounts need leaving alone, which is worth saying because customer success capacity is finite and it tends to get spent on whoever complained most recently.For definitions see churn prediction and customer health score. For how retention is measured once churn happens, see the ARR waterfall that reconciles.
Frequently Asked Questions
Does a customer with no support tickets mean they are happy?
No. A customer with no support cases is at risk of churn. Silence usually means low usage or low engagement rather than a product working perfectly, and a customer who never needs help is often a customer who is not using the product deeply enough to keep paying for it.
How many support tickets indicate a healthy customer?
Three to five tickets, typically tier 2 or tier 3 rather than severe, is the healthiest range. Those customers are engaged, they are getting support, and they are usually happy. Seven or more in a year moves back into risk territory.
Why is support ticket volume a U-shaped churn signal?
Because both ends represent failure modes. Zero tickets indicates disengagement, and a high volume indicates unresolved friction. The middle of the range indicates a customer using the product enough to hit questions and getting them answered.
Why does silence remove the renewal case?
Because support interactions are contact points where value gets demonstrated and where someone at the vendor learns what the customer is trying to do. A customer with zero tickets has had no such contact, so at renewal there is no accumulated evidence of value.
Does ticket severity matter as well as volume?
Yes. Three to five tier 2 or tier 3 tickets look like a customer working through normal questions. A comparable count of severe tickets is a different situation entirely, and volume alone will not distinguish them.
See how ORM turns these insights into action
ORM builds custom revenue forecast models for B2B SaaS companies. Not dashboards. Prescriptive analytics that tell you what to do next.
Schedule a Demo