CHURN June 2026 · 7 min read

The 7 Churn Signals Every SaaS Founder Ignores

They're not hidden. They're not subtle. They're sitting in your data right now, and most founders walk past them every single day. Here's what they look like — and why missing them is costing you customers you could have kept.

Churn is the metric everyone tracks and almost nobody predicts. The typical SaaS playbook goes: customer signs up, uses the product for a while, sends a cancellation email, gets a discount offer, leaves anyway. Repeat.

What makes this frustrating is that the warning signs were almost certainly there. The problem is that most founders are looking for churn after the fact — analyzing cancelled accounts — instead of monitoring active accounts for the signals that precede cancellation.

Here are the 7 most commonly ignored signals, what they actually mean, and what to do when you see them.

The 7 signals

SIGNAL 01 · HIGH RISK
The "polite ghost"
Your champion replies to emails but with one-word answers. "Sure." "Thanks." "Got it." They used to send paragraphs. This isn't them being busy — it's them mentally stepping back from the relationship. Polite ghosts churn within 60 days in the majority of cases.
SIGNAL 02 · HIGH RISK
The skipped QBR
A customer who cancels a quarterly business review without rescheduling has silently deprioritized your relationship. One skip is a calendar conflict. Two skips is a signal. Three skips means the relationship is already over — you just haven't gotten the email yet.
SIGNAL 03 · HIGH RISK
The "just checking" contract request
When a customer asks for a copy of their contract "just to review the terms," they are evaluating whether to renew. Nobody reviews contracts out of idle curiosity. This is one of the clearest pre-churn signals in B2B SaaS and almost everyone treats it as an admin request.
SIGNAL 04 · MEDIUM RISK
The power user goes quiet
Every account has one person who actually uses the product heavily. When their activity drops — fewer logins, shorter sessions, fewer features used — the account is hollowing out from the inside. This matters more than aggregate account-level usage numbers.
SIGNAL 05 · MEDIUM RISK
The feature request that stops
Engaged customers ask for things. They submit feedback, vote on roadmap items, request integrations. When an account that used to be vocal about feature requests goes completely quiet, they've stopped investing in the product's future — because they're not planning to be around for it.
SIGNAL 06 · HIGH RISK
The new stakeholder who wasn't introduced
If you find out through a CC on an email that there's a new VP or director overseeing your account — and nobody introduced you — your champion didn't advocate for you. Their new boss has no relationship with you and no reason to keep paying for a tool they didn't choose.
SIGNAL 07 · MEDIUM RISK
The G2 visit spike
When traffic from review sites like G2 or Capterra spikes to your site — especially to your pricing page — it often means a customer is benchmarking you against alternatives. They're not researching you because they love you. They're building a business case to switch.

Why founders miss these signals

It comes down to three structural problems:

1. The data is scattered

Signal 1 lives in your email. Signal 3 lives in your CRM. Signal 4 lives in your product analytics. Signal 6 lives in LinkedIn. No single tool surfaces all of them together, so nobody connects the dots until it's too late.

2. Optimism bias

Founders and CS teams are naturally optimistic about accounts. The polite ghost feels like a busy customer, not a pre-churn signal. The skipped QBR gets attributed to scheduling. This isn't incompetence — it's human nature. But it costs you accounts.

3. No scoring system

Without a defined risk score, every account looks roughly the same until it doesn't. The solution is to assign weight to each signal and automatically flag accounts when they cross a threshold — not rely on someone noticing something feels off.

The compound effect One signal in isolation is ambiguous. Three signals in 30 days is a pattern. The accounts most likely to churn almost always show multiple signals simultaneously — which is exactly why automated scoring beats manual monitoring every time.

Building a simple churn signal playbook

You don't need a complex machine learning model to start catching these signals. You need a defined process:

  1. Assign risk weights — high-risk signals (ghost, contract request, skipped QBR) score 3 points each. Medium-risk signals score 1 point each.
  2. Set a threshold — any account scoring 4+ points in a rolling 30-day window triggers an alert.
  3. Define a response playbook — what happens at 4 points (personal outreach), what happens at 7+ points (escalate to founder).
  4. Review weekly — dedicate 20 minutes every Monday to reviewing your at-risk accounts list and confirming the right actions are in motion.

The goal isn't to save every at-risk account. Some will churn regardless. The goal is to make sure you had the conversation — that you saw the signal, responded intentionally, and gave the relationship a real chance before the cancellation email arrived.

Stop missing churn signals across your accounts

Signal Engine monitors your accounts for all 7 of these signals automatically — and flags risk before you'd catch it manually. Free 7-day trial.

Start Free Trial →
Free 7-day trial · No credit card · Cancel anytime
BD
Bernard Downing
Founder, Signal Engine
← All articles