Onboarding Playbook

No onboarding playbook exists in any form. Customer success managers wing every new customer activation based on memory and instinct. 'Every onboarding is different because we're basically reinventing it each time.' When a CSM leaves, all their onboarding knowledge walks out the door with them. New CSMs shadow for weeks and still miss critical steps.

None — AI cannot guide or optimize customer onboarding because no onboarding playbook exists in any system to reference, sequence, or measure against.

A basic onboarding playbook exists as a Google Doc or wiki page that someone wrote months ago. It lists general steps like 'schedule kickoff call' and 'configure integrations' but lacks specific task owners, duration estimates, or success criteria. Each CSM interprets the playbook differently. 'The playbook says do a health check but doesn't define what that means or when it's done.'

AI can display the onboarding playbook as a reference document, but cannot track progress, identify delays, or personalize the onboarding sequence because the playbook lacks structured milestones and measurable completion criteria.

The onboarding playbook has structured milestones with defined tasks, owners, and expected timelines. Each milestone has clear completion criteria — 'SSO configured and tested with 3+ users' rather than 'set up SSO.' Playbooks are stored in the CS platform with consistent fields. But playbooks are one-size-fits-all with no variation by customer segment, product tier, or use case.

AI can track onboarding playbook progress against milestones, flag overdue tasks, and generate status reports. Cannot personalize onboarding sequences because the playbook has no segment-specific variations or conditional branching logic.

Onboarding playbooks are connected knowledge assets with segment-specific variants. An enterprise healthcare customer gets a different milestone sequence than a mid-market retail customer. Each playbook links to related customer records, product configuration templates, and training content libraries. A CSM can query 'show me the standard onboarding playbook for enterprise customers purchasing our analytics module' and get the specific, contextualized sequence.

AI can recommend the optimal onboarding playbook variant for a new customer, predict time-to-value based on historical playbook completion patterns, and flag at-risk onboardings by comparing progress against segment benchmarks. Cannot dynamically restructure the playbook mid-flight based on real-time customer behavior.

The onboarding playbook is a formal workflow ontology with typed milestone entities, dependency chains, conditional branching rules, and quantified success criteria. An AI agent can reason about the playbook structure: 'this customer skipped the integration milestone — based on the dependency graph, training module 3 won't be effective until integration is complete. Recommend rescheduling.' Every playbook element is a queryable, structured object.

AI can autonomously orchestrate standard onboarding sequences, dynamically reorder milestones based on customer progress, generate personalized task lists, and predict completion dates with high accuracy. Human judgment needed for novel customer situations that fall outside established playbook patterns.

The onboarding playbook is a self-evolving workflow model. Milestone definitions, sequencing rules, and success criteria continuously update based on observed customer activation patterns. When product features change, the playbook auto-adjusts its task definitions. New best practices discovered through successful onboardings are automatically codified into the playbook structure. The playbook writes itself from outcomes.

Can autonomously design, execute, and optimize onboarding playbooks for any customer segment. The playbook evolves its own structure from real-time outcome signals, requiring no manual maintenance or periodic review.

Ceiling of the CMC framework for this dimension.