Ghosted Account Re-Engagement Playbook 2026
Ban the "just checking in" email and replace it with the customer's own data: 30 days of silence plus a declining engagement score fires the play carrying the score trend, the last meaningful action and the features gone quiet — so every touch, from the value-led email to the AM's insight call to the physical send for high-ARR accounts, opens with something they didn't know about their own usage.
"Just checking in" is CS spam, and it should be banned from your team's vocabulary. It transfers no information, asks the customer to do your job of finding something to talk about, and trains them to delete your emails — which is exactly why ghosted accounts stay ghosted. No reply is not no signal: the product data keeps talking the whole time. This play makes a rule of it: no touch goes out without an insight from the customer's own usage. Thirty days of human silence plus a declining engagement score fires the play carrying its own conversation starters — the score trend, the last meaningful action they took, the features that went quiet — and the only decision is which format the insight travels in: an automated email while usage is alive, an AM call when it's dying, a physical send when the account is worth more than an inbox.
Measure it on the reply rate on re-engagement touches, meetings booked from ghosted accounts, and the number of accounts sitting at 30+ days without human contact — the queue this play exists to drain.
How it works7 steps
01SignalDetect the ghost with two conditions, not one
Silence alone isn't a trigger — plenty of healthy accounts don't need to talk to you, and interrupting them costs goodwill. The trigger is the intersection: no meaningful human contact in 30 days (email replies and meetings count; your outbound doesn't) AND a declining engagement score. Accoil computes the score from the events you already send via Segment, PostHog, Amplitude or Mixpanel, and the trigger fires carrying the payload every later touch is built from: the score trend, the last meaningful action the account took, which features went quiet, and the date of the last real contact. That payload is the whole point — nobody downstream should have to go find something to say.
02DecisionRoute on whether the product is still doing its job
One question splits the play: is usage still alive? A quiet relationship with healthy usage is a communication problem — automate it. A quiet relationship with dying usage is a churn risk wearing headphones — that gets a human. Set the line mechanically: score trend flat-or-up with active seats holding means alive; a double-digit score slide over 30 days means dying. No judgment calls at triage time.
03ActionSend the email only their data could write
For alive-but-silent accounts, send one email keyed to their actual usage: "your team shipped 40% more X last month — here's what teams like yours do next." The numbers come from the payload; Customer.io just delivers them.
- One insight, one next step, one sentence of each. If the email works without the merge fields, it's not this email — delete it and start over.
- Point the next step at a quiet feature adjacent to what they already use; adoption grows along edges, not leaps.
- Cap it at one per month per account, and suppress the sequence the moment a human conversation starts. Automation that keeps talking over your AM is worse than silence.
04ActionPut the dying account in front of a human
Dying usage doesn't get an email — it gets flagged to the account owner in #cs-alerts with the full payload: score trend, last meaningful action, the features that went dark, days since last contact. The AM should be able to dial from the alert without opening a dashboard. Shared channel over DM, always: unclaimed at-risk accounts need to be visible to someone whose job is noticing.
05Human stepAM makes the insight call, never the check-in call
The call opens with what you observed, not how they're doing: "your team's usage of the reporting module dropped by half since March — did the workflow move somewhere else, or did something break?" Then stop talking. The insight does two jobs the check-in can't: it proves you're paying attention, and it gives the customer something concrete to correct, confirm or explain — any of which restarts the relationship. Log the reason in the CRM with a code; a quarter of these calls tells you why accounts actually go quiet, which is a churn-driver report most teams would pay for.
06ActionEscalate high-ARR silence to a physical touch
When a top-tier account stays silent after the call attempt, go physical — but keep the rule: the send references their usage, not your logo. A note that says "your team hit 10,000 projects in our product last quarter — congratulations from ours" lands because it's unfakeable attention. Gate it hard: reserve Sendoso sends for your top ARR tier or accounts above your save-motion threshold, and skip this step entirely when the insight call landed. A gift that arrives mid-conversation reads as a bribe; one that arrives during silence reads as effort.
07OutcomeExit on contact or recovery, and count it
The play closes one of two ways: a meeting booked, or the engagement score back above its trailing baseline with the silence broken. If neither happens within 45 days of the first touch, stop touching — route the account to your churn-risk save motion instead, because at that point you're not re-engaging, you're triaging. Every exit gets logged against the touch that produced it, so you learn which format — email, call, send — actually moves your book.
The debate
The strongest objection comes from the automation-at-scale camp, and it's fair: personalization doesn't scale. A CSM carrying 150 accounts cannot hand-craft insights for every quiet logo, and a scheduled check-in — however empty — at least keeps the relationship warm and the vendor's name in the inbox. An imperfect touch that ships beats a perfect touch that doesn't, and banning the check-in mostly means fewer touches.
Our answer: a check-in without information isn't a warm touch, it's a notification that you have nothing to say — and customers learn to treat it exactly like the spam it resembles, which makes the next touch weaker too. The scale argument also quietly assumes the insight is expensive. It isn't: the payload arrives pre-assembled — score trend, quiet features, last action — so the data-led email is automated end-to-end and the insight call takes the same five minutes as the check-in call. Same effort, but one of them transfers value.
How Accoil fits
Accoil is the layer that makes the no-empty-touches rule affordable: it watches the events you already collect, notices the silence-plus-decline pattern, and fires the trigger with the conversation pre-loaded — trend, last meaningful action, features gone quiet. Customer.io delivers the data-led email, Slack routes the dying accounts to a human, Sendoso carries the high-ARR escalation. Every touch in the ladder opens with something the customer didn't know about their own usage, and none of it requires anyone to go looking for it.
Tools are interchangeable here — run the emails from Ortto or Userlist instead of Customer.io, read the alerts in Teams instead of Slack — and the ladder holds, because Accoil pushes the same signal into whichever tools do the work.
Accoil is the scoring layer in this playbook — it works on the product events you already collect, and shows your accounts scored in under 48 hours. Free to start, no credit card.
Explore Accoil →Keep reading
Monday-Morning Account Triage Playbook 2026
Most CSM weeks are planned by the inbox: whoever emailed last gets the attention. This play starts the week from the product data instead — a Monday digest of the biggest score moves, stalled onboardings and upcoming renewals, triaged in fifteen minutes into three piles: save touches to make now, expansion signals for sales, and thriving accounts to leave alone.
The NPS Autopsy Playbook 2026
Kill the quarterly NPS blast — survey behavior instead. When an account crosses a behavioral threshold (a sustained spike, a 20-point decline, an activation milestone), the moment picks the micro-survey: power users get the advocacy ask, decliners the friction probe, the newly-activated value confirmation. Every answer is joined with the respondent's actual usage and routed to a different play.
AI Agent Account Context Playbook 2026
AI agents are only as useful as what they know about the account. Give yours the whole picture — engagement and usage from Accoil, contract and pipeline from the CRM, billing from Stripe, support history, session replays — and every signal comes back as a one-page brief, a drafted outreach and a suggested next step. The CSM reviews and sends; the judgment stays human, the hour of prep doesn't.
Every playbook, one download
All 31 workflows as print-ready playbooks — diagrams included. Plus every new workflow as we publish it.