Hank watches your customers for the signs of leaving.
Hank watches your customers for the signs of leaving. He flags accounts that have gone quiet (no logins, no API calls, no playbook runs in 60+ days) before they cancel, and surfaces the patterns — feature drop-off, support-ticket spikes, login frequency declines — that predict churn. The signals are early enough to do something about. The framing is honest, not alarmist.
What Hank does.
Silent churn is the worst kind. The customer doesn't complain, doesn't email, doesn't ask for a refund — they just stop using the product, and three months later they cancel without warning. By then, the relationship is over and there's no save play.
Hank watches for the early signals: a customer who logged in daily for six months suddenly going silent, a customer whose API call volume drops 70%, a customer who stops triggering playbooks they used to rely on. Each gets a flag with the data behind it and an honest summary of what's changed.
Hank doesn't try to save every flagged account. Some are off — customers do go on vacation, customers do build their own internal tools, customers do graduate to a bigger platform. Hank's job is to surface the signal so the founder can decide whether to reach out. That decision belongs to the founder.
What sets it off, what it reads, what it writes.
- Triggers
- Daily silent-customer scan (no activity in 60+ days; configurable)
- Usage anomaly detected (rolling 7-day vs 90-day baseline)
- Support ticket spike on a single account
- Login frequency decline (50%+ drop sustained over 14 days)
- Inputs
- Customer activity log (logins, API calls, playbook runs)
- Customer-graph history (tenure, MRR, prior touchpoints)
- Support ticket history
- Lifecycle stage (new, active, late)
- Outputs
- Quiet-customer flag with rationale
- Usage-trend summary card
- Suggested followup template (drafted, founder reviews)
- Audit trail entry
The playbook in plain English.
Every day, Hank runs the silent-customer scan: customers whose last meaningful activity is 60+ days old (configurable) get reviewed. He pulls the full customer-graph history — when they joined, what they used to do, what changed — and writes a one-paragraph summary that goes on the founder's morning brief.
In parallel, Hank watches for usage anomalies in active customers. A customer whose API call volume drops 70% over a rolling 14-day window gets flagged with a usage-trend chart. A customer who's stopped triggering a key playbook gets flagged. Hank doesn't predict churn; he names what's changed and lets the founder interpret.
When Hank flags a customer, he drafts a no-pressure followup the founder can send: "noticed you haven't logged in much lately — anything blocking you we can help with?" The framing is genuinely curious, never sales-y. The founder edits, sends, or marks the customer as "intentional pause" so Hank stops watching that account temporarily.
Hank doesn't generate alarm. A flagged customer isn't a churning customer; they're a customer worth a conversation. Six months in, the audit log tells you which Hank flags were saves, which were customers who were just busy, and which were genuinely on the way out. That feedback loop tightens his thresholds.
The guardrails are yours.
- Silent threshold — days of inactivity before Hank flags (default 60).
- Usage anomaly threshold — percent drop over rolling window that triggers a flag (default 70% over 14 days).
- Suppression list — customers who should never be flagged (intentional pauses, integration testing, internal accounts).
- Followup auto-send — auto-send the "noticed you've been quiet" note (default off; founder always reviews).
- Daily flag cap — max customers Hank flags per day (default 5; you can act on, not 50).
- Tone profile — Hank's followup drafts match the founder's voice samples.
What Hank has shipped for porchops.
- 7
Quiet customers Hank flagged for porchops in March; 3 turned into save conversations.
- 60 days
Silent threshold (down from 90 once the false-positive rate stabilized).
- 0
Cancellations from accounts Hank had flagged 30+ days prior.
- 84%
Of Hank's silent-customer flags result in a meaningful founder conversation.
Real quotes.
Hank flagged Bramble at day 65 of silence. I sent the followup, turned out their API integration broke during their migration to a new stack and they didn't tell us. Saved.
Travis · cofounder, Porchops
Hank on the Customers Post.
Hank is the watcher Op on the Customers Post. He completes the customer journey loop: Maggie activates new users, Cleo handles renewals and expansion, Hank watches for silent churn. Together they cover what happens after acquisition.
Read the customers canon →Last few times the playbook ran.
- apr 27Telos · 90-day silent · followup drafted · awaiting Travis review
- apr 24Bramble · API call volume dropped 75% · flagged · followup queued
- apr 19Heliotrope · login frequency drop · flagged · followup sent (saved)
Common questions about Hank.
Does Hank predict churn?
No. Hank surfaces signals — silent-customer thresholds, usage anomalies, login-frequency declines. The founder decides whether each flag is a churn risk. Predicting churn is a different problem; surfacing the signal is achievable, useful, and honest.
Why 60 days for silent?
Empirically, 60 days is the sweet spot for solo SaaS at our scale. 30 days has too many false positives (customers go on vacation); 90 days catches issues too late to save. You can configure the threshold per cohort.
What does Hank do for active customers?
He watches for usage anomalies — sudden drops, sustained declines, support-ticket spikes on a single account. Active customers don't trigger the silent-customer flag, but they do trigger anomaly flags when something materially changes in their usage.
How do you avoid alarming false positives?
Conservative thresholds, daily flag cap (default 5), and suppression list for known intentional pauses. The audit log records every flag and outcome; thresholds tighten as feedback accumulates. Six weeks in, false-positive rate stabilizes around 15%.
What's the difference between Hank and Cleo?
Cleo handles the calendar — renewal preparation, expansion timing. Hank handles the unscheduled — silent customers, usage anomalies. They're complementary: Cleo is the planned touchpoints; Hank is the response to unexpected signals.
What's the followup tone?
Genuinely curious, never sales-y. The default draft is "noticed you've been quiet — anything blocking you we can help with?" — short, no pressure, no "upgrade" pitch. The founder edits if needed; the tone is supposed to feel like a friend checking in.
Closing
Hank names what's slipping. The save play is yours. — Travis