skills/customer-success/SKILL.md
Use when managing post-sale customer relationships, building health scores, predicting churn, preparing QBRs, designing renewal playbooks, identifying expansion revenue, creating customer segmentation models, building NPS/CSAT programs, or designing escalation frameworks. Use when analyzing customer lifecycle stages from onboarding through advocacy. Do NOT use for lead generation, prospecting, or pre-sale activities (use lead-research-assistant). Do NOT use for deal closing, proposal writing, or sales pipeline management (use sales-enablement). Do NOT use for product feature development or roadmap planning.
npx skillsauth add sharkitect-solutions/sharkitect-claude-toolkit customer-successInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
| File | Load When | Do NOT Load |
|---|---|---|
| references/health-scoring-framework.md | Building or refining health scores, defining thresholds, setting up automation rules, benchmarking account health | General renewal conversations, QBR prep without scoring focus |
| references/qbr-playbook.md | Preparing quarterly business reviews, structuring executive presentations, defining QBR cadence, measuring QBR effectiveness | Health scoring without review context, churn analysis |
| references/churn-signals-library.md | Analyzing churn risk, building early warning systems, designing intervention playbooks, conducting churn post-mortems | Healthy account expansion planning, new customer onboarding |
| Domain | This Skill Covers | Use Instead | |---|---|---| | Pre-sale leads | -- | lead-research-assistant | | Deal closing / proposals | -- | sales-enablement | | Product roadmap decisions | Customer feedback synthesis only | Product management | | Support ticket resolution | Escalation framework + trends | Support engineering | | Billing disputes | Churn signal detection only | Finance / billing ops | | Marketing campaigns | Advocacy program referrals only | marketing-demand-acquisition |
Define every account's current stage. Stage determines which playbooks apply and which metrics matter.
Never skip stages. An account in Onboarding that requests expansion still completes onboarding milestones first. Skipping stages creates fragile relationships that collapse at renewal. Why: incomplete onboarding is the #1 predictor of first-year churn across all SaaS cohort studies.
Health scores are multi-dimensional. A single composite number hides critical risk signals. Track five dimensions independently, then combine.
Use the weighted composite with floor override method:
When data points go missing for 30+ days, decay that dimension by 5 points per missing month. Why: absence of data is itself a signal -- disengaged customers stop generating telemetry before they stop paying.
Detect these early and intervene before the customer mentally commits to leaving:
These indicate the decision is likely made. Shift from prevention to win-back:
Prioritize leading indicators in all dashboards and alerts. Lagging indicators are useful for post-mortem accuracy, not prevention. Why: by the time lagging indicators fire, the customer has already evaluated alternatives.
T-90 days (Assessment):
T-60 days (Alignment):
T-30 days (Execution):
T-7 days (Close):
| Failing Dimension | Intervention | |---|---| | Usage declining | Conduct re-onboarding sprint, assign adoption specialist | | Support health red | Executive apology + dedicated support engineer for 30 days | | Financial stress | Offer bridge pricing, defer expansion, reduce scope | | Relationship gaps | Multi-thread aggressively, secure new executive sponsor | | Onboarding stalled | Reset onboarding plan, assign implementation PM |
Expansion revenue compounds retention value. A 110% net revenue retention means the business grows even with some churn.
Never pitch expansion to a Yellow or Red account. Fix health first. Why: expansion attempts on unhealthy accounts accelerate churn by signaling you prioritize revenue over their success. The sole exception: when expansion directly solves the health problem (e.g., upgrading support tier for a support-health-red account).
| Segment | ACV Range | CS Model | Touch Frequency | |---|---|---|---| | Enterprise | >$100K | Named CSM, dedicated | Weekly | | Mid-Market | $25K-$100K | Named CSM, pooled (1:25) | Bi-weekly | | SMB | $5K-$25K | Pooled CSM (1:75) | Monthly | | Self-Serve | <$5K | Tech-touch only | Automated |
| Health | Action Model | |---|---| | Green 80+ | Expansion-focused, advocacy programs, case study candidates | | Yellow 60-79 | Proactive monitoring, bi-weekly check-ins, risk remediation | | Red 0-59 | Intervention mode, executive escalation, save playbook activated |
Apply stage-appropriate playbooks. An Enterprise account in Onboarding gets different treatment than an Enterprise account in Renewal. Segment by value first (determines resource allocation), then by health (determines urgency), then by stage (determines playbook).
| Level | Criteria | Response SLA | Owner | |---|---|---|---| | SEV-1 | Revenue at risk >$100K, exec sponsor escalation, legal threat | 4 hours | VP of CS | | SEV-2 | Health score Red for 30+ days, renewal at risk, multi-issue compound | 24 hours | CS Director | | SEV-3 | Single dimension Red, feature blocker, support dissatisfaction | 48 hours | Senior CSM | | SEV-4 | Yellow trending down, minor dissatisfaction, process complaint | 1 week | Assigned CSM |
Never escalate without a recommended action. Escalations without recommendations waste leadership cycles and delay intervention. Why: the CSM closest to the account has the best context for initial remediation strategy.
| Score Range | Classification | Required Action | |---|---|---| | NPS 9-10 | Promoter | Invite to advocacy program, request referral, case study candidate | | NPS 7-8 | Passive | Investigate barriers to delight, assign improvement plan | | NPS 0-6 | Detractor | Trigger escalation framework SEV-3 minimum, personal outreach within 48h |
Close the loop on every detractor response within 5 business days. Why: detractors who receive follow-up are 3x more likely to remain customers than detractors who are ignored (Bain & Company research).
| Situation | Wrong Approach | Right Approach | Why | |---|---|---|---| | Champion leaves customer org | Continue emailing departed contact | Immediately multi-thread: identify 3+ new contacts across departments within 2 weeks | Single-threaded relationships are the #1 controllable churn factor | | NPS detractor response received | Log it in dashboard, review next quarter | Personal outreach within 48 hours, escalate to SEV-3 | Speed of response to negative feedback directly correlates with save rate | | Usage drops 25% month-over-month | Wait to see if it recovers next month | Trigger proactive outreach within 72 hours, investigate root cause | Usage recovery without intervention occurs <15% of the time | | Customer requests discount at renewal | Grant discount to save the deal | Quantify delivered ROI first, negotiate on scope not price | Discount dependency creates downward revenue spiral and trains customers to threaten churn | | Account health is Green across all dimensions | Reduce touch frequency to focus on Red accounts | Maintain cadence and pivot to expansion/advocacy motions | Neglecting healthy accounts turns Green into Yellow within 2 quarters |
Assuming silence equals satisfaction. Disengaged customers stop complaining before they stop paying. Silence is the most dangerous churn signal because it generates no alerts in reactive systems. Require proactive health checks regardless of inbound signal volume.
Celebrating customer count retention while ignoring net revenue retention. Keeping 95% of logos means nothing if the retained accounts all downgraded. Track gross revenue retention (without expansion) and net revenue retention (with expansion) separately. The logo number goes on the slide deck; the revenue number drives strategy.
Presenting a 45-minute product demo, calling it a Quarterly Business Review, and wondering why executives stop attending. QBRs exist to align on customer goals and demonstrate value delivered against those goals. If the customer talks less than 40% of the meeting, it is a sales pitch, not a QBR.
Only engaging with accounts after they turn Red. By the time a health score hits Red, the customer has been dissatisfied for weeks or months. Proactive CS means running expansion plays on Green accounts and remediation plays on Yellow accounts -- not waiting for fires.
Treating every feature request as a churn risk and escalating to product. Most feature requests are aspirational, not existential. Distinguish between "nice to have" requests and "I will leave without this" blockers by asking: "If we never build this, what happens to your workflow?" The answer separates noise from signal.
development
When the user wants help with paid advertising campaigns on Google Ads, Meta (Facebook/Instagram), LinkedIn, Twitter/X, or other ad platforms. Also use when the user mentions 'PPC,' 'paid media,' 'ad copy,' 'ad creative,' 'ROAS,' 'CPA,' 'ad campaign,' 'retargeting,' or 'audience targeting.' This skill covers campaign strategy, ad creation, audience targeting, and optimization.
testing
--- name: using-sharkitect-methodology description: Use when starting any conversation in a Sharkitect workspace OR before any task involving NEW pricing, positioning, proposal, strategy, plan-execution, or schema-design work — mandates invocation of Sharkitect-specific methodology skills (pricing-strategy, marketing-strategy-pmm, smb-cfo, hq-revenue-ops, executing-plans, brainstorming) under the same anti-rationalization discipline as using-superpowers. Documentation has failed 4 times across H
testing
Use when user says 'end session', 'wrap up', 'stop for the day', 'done for today', 'close out', 'save session', 'wrapping up', or invokes /end-session. Runs the full 9-step end-of-session protocol: resource audit, MEMORY.md update, lessons capture, plan status, pending items, workspace checklist, .tmp/ audit, git commit+push, Supabase brain sync, session brief, summary. Final step schedules a detached self-kill of the current session ONLY (3s delay) so the window closes cleanly. Other claude.exe processes (active workspaces) are NOT touched -- orphan cleanup is handled separately by Claude-Orphan-Cleanup-Hourly with proper age safeguards. Do NOT use for: mid-session quick saves (use session-checkpoint), skill syncing (use sync-skills.py), brain memory queries (use supabase-sync.py pull), document freshness reviews (use document-lifecycle), resource gap detection (use resource-auditor).
testing
Remove signs of AI-generated writing from text. Use when editing or reviewing text to make it sound more natural and human-written. Based on Wikipedia's comprehensive "Signs of AI writing" guide. Detects and fixes patterns including: inflated symbolism, promotional language, superficial -ing analyses, vague attributions, em dash overuse, rule of three, AI vocabulary words, passive voice, negative parallelisms, and filler phrases.