If you manage customer retention, you know the pain: A user cancels and it blindsides your team. You probably already track some health scores or NPS, but what about those subtle red flags that slip through? That’s where automating churn risk alerts can actually make a difference—if you set it up right.
This guide is for folks using Custify who want to catch churn risks early, using custom triggers based on real warning signs—not just generic, “usage is down” alerts. If you want practical steps and no-nonsense advice, you’re in the right place.
Why bother with custom churn risk alerts?
Let’s be blunt: most out-of-the-box churn alerts are too generic. “Low activity” or “no login in 7 days” is helpful, but it misses a lot of context. You end up with too many false positives, or worse, you tune it out entirely.
Custom triggers let you:
- Flag the right risks for your business, not someone else’s
- Spot problems before they snowball (think: key decision-maker goes silent, support tickets spike, or payment fails)
- Stop wasting your team’s time on noise
But, don’t expect miracles. No automation is perfect. It’s about stacking the odds in your favor—catching churn early is a process, not a silver bullet.
Step 1: Get clear on what “churn risk” actually means for you
Before you touch Custify, decide what really signals danger for your users. This is where most teams mess up—they just copy standard triggers and end up with alert fatigue.
Ask yourself:
- What’s happened just before a customer churned in the past? (Dig into your own data, not just industry blog posts.)
- Are there behaviors that always lead to churn, or is it more subtle? (E.g., multiple support tickets + CSM ignored = red flag.)
- Who on your team needs to know, and what do you want them to do when an alert fires?
Common churn signals that are actually useful: - Key user stops logging in (especially if they’re the contract signer) - Sudden drop in product usage, after a good onboarding - Repeated negative NPS or CSAT scores - Unresolved support issues piling up - Payment failures or subscription downgrades
Stuff that rarely works:
- “Hasn’t logged in in 3 days” (unless your product is daily-use)
- Activity dip during a known holiday or seasonal slowdown
- Alerts based on only one metric
Don’t try to catch everything. Start with 2–3 triggers that match your churn patterns.
Step 2: Make sure your data is flowing into Custify
You can’t trigger alerts without real data. Before you set up anything, double-check:
- Product usage data: Are logins, feature use, and key events tracked and mapped to the right accounts?
- Support integrations: Is your help desk (Zendesk, Intercom, etc.) connected so you can see ticket volume?
- Billing info: Can you see failed payments, downgrades, or upcoming renewals in Custify?
- Custom fields: Need to flag a ‘decision maker’ or a ‘VIP’ client? Make sure those are set up.
Pro tip: Don’t trust that your data is “probably fine.” Test your integrations by looking up a few accounts and checking if the key fields update in real time.
If something’s missing, fix it now—otherwise, you’ll be flying blind and your alerts will be garbage.
Step 3: Define your custom churn risk triggers
Now, time to get hands-on in Custify. Don’t worry, you don’t need to be a developer, but you do need to think logically.
3.1. Go to the Automation (Playbooks) section
- In Custify, head to the “Playbooks” or “Automations” tab.
- Click “Create New Playbook” (or similar—Custify tweaks UI labels now and then).
3.2. Set up your trigger conditions
- Choose your object: Account or User (most churn risk alerts are at the Account level)
- Pick your trigger: “When [event/condition] happens…”
Example triggers you might actually want: - “Account’s main user hasn’t logged in for 14 days” - “More than 3 unresolved support tickets in 7 days” - “NPS dropped below 6 twice in a row” - “Payment failed in the last billing cycle”
You can combine conditions using AND/OR logic, but don’t get too clever. Complex logic gets hard to debug and maintain.
What doesn’t work:
Setting up a dozen triggers for every possible scenario. You’ll overwhelm your team and yourself.
3.3. Set frequency and suppression rules
- Decide if alerts should fire every time, or only once per account per X days.
- Suppress duplicates so your team doesn’t get spammed.
Step 4: Choose your alert actions
Once a trigger condition is met, what happens? Custify lets you automate several things. Here’s what’s actually useful:
- Send an email or Slack alert to the account owner (fastest way to get someone’s attention)
- Create a task in Custify for a CSM (ensures follow-up)
- Update a custom field (like ‘Churn Risk = True’) for reporting
- Push the alert to your CRM if you want sales or execs looped in
Be wary of: - Automated templated emails to customers—these can come off as robotic and insincere. Better to nudge your team to reach out personally.
Set up just 1–2 alerting methods to start. You can always add more once you see how it works.
Step 5: Test your triggers with real data
Before you roll this out for all customers, test on a few accounts:
- Use real (but non-critical) accounts to trip your triggers. Adjust your conditions if you’re getting too many or too few alerts.
- Check that data updates fast enough—delays kill the usefulness of alerts.
- Make sure your team knows what the alert means and what they’re expected to do.
Pro tip: Don’t assume “it works” because you got the alert. Ask your team if the alert helped, or just added noise.
Step 6: Monitor and fine-tune
No matter how clever your triggers, some will be duds. That’s normal. Plan to revisit your alerts every few weeks:
- Track: How many alerts led to real saves vs. false alarms?
- Ask your team: Are alerts actionable? Are they getting ignored?
- Tweak your conditions—loosen or tighten as needed.
If you notice alert fatigue, kill or consolidate triggers. It’s better to have 1 good alert than 10 bad ones.
Honest advice: What to avoid
- Don’t automate everything. Some things need human judgment. Use triggers as a heads-up, not a replacement for talking to your customers.
- Don’t ignore context. Seasonality, account type, or relationship history often matter more than a single metric.
- Don’t overcomplicate. The more you automate, the harder it gets to know what’s actually working.
- Don’t forget to act. The best alert is useless if nobody does anything.
Wrapping up: Keep it simple, iterate often
Churn risk automation isn’t magic, but it is useful if you keep it grounded. Start with a couple of custom triggers that reflect real-world churn signals for your business. Make sure your data is solid, test before rolling out, and resist the urge to automate every possible scenario.
The best setup is the one your team will actually use. You can always add more triggers as you learn what works. Stick to what’s actionable, keep alerts human, and you’ll catch churn risks before they become cancellations. That’s as good as automation gets.