Build Your CS Playbook in 90 Days – Kerin Smollen

A complete CS playbook at Series A includes five components: an onboarding sequence with day-by-day milestones from contract signature through first value moment – typically day 1 through day 45; a health scoring model with 5–8 metrics derived from your own usage data, weighted by their predictive value for renewal versus churn; a QBR template structured around the customer's business outcomes, not your product's feature list; an escalation protocol that tells every customer-facing person exactly what to do when a customer shows risk signals – who owns the escalation, what the response timeline is, and what the recovery playbook looks like; and a renewal playbook that begins 90 days before contract end. Most companies have a rough version of the first and nothing else.

Hire first when you have enough customers that not having a dedicated CS person is actively causing damage and the founding team genuinely doesn't have capacity to build the playbook. Build the playbook first when you can sustain customer relationships for another 60–90 days and want to set your first CSM hire up to execute rather than build infrastructure. The cost comparison is straightforward: a CS playbook build engagement costs a fraction of one CSM annual salary, and it makes every subsequent CSM hire 30–50% more effective from their first month. Companies that hire first and build later report that the CSM spent their first 90 days doing documentation work instead of customer work – which is the most expensive way to build a playbook.

Health scoring without historical data starts with a hypothesis model rather than a statistical one. Identify 5–8 candidate metrics – typically a combination of product usage frequency, feature adoption breadth, support ticket volume, and stakeholder engagement signals – and assign weights based on the team's collective judgment about what drives customer success. Then run that model against your current customer base and validate whether the scores match your intuitive assessment of who is healthy and who is at risk. The model gets calibrated over the first two renewal cycles as you see which metrics actually predicted churn. Starting with an imperfect hypothesis model is always better than waiting for enough historical data to build a statistically validated one.

The engagement produces five concrete deliverables: a documented onboarding sequence with milestone definitions and owner assignments for each step; a calibrated health scoring model implemented in your CRM or CS tool; a QBR template with guidance on how to run the conversation; an escalation protocol with a decision tree and owner matrix; and a renewal playbook with a 90-day countdown checklist. The engagement also includes a working session with the founding team to validate the onboarding sequence against real customer experiences, and a handoff session so the first CSM hire can get up to speed in one week rather than three months.

Three interventions have the highest impact before a dedicated CS team is in place. First, standardize the onboarding handoff from sales – most churn begins at the moment when a new customer completes the sales process and nobody owns their success. A documented onboarding checklist and a single point of contact for the first 45 days closes most of that gap. Second, identify your top five at-risk customers using whatever data you have – usage logs, support tickets, stakeholder silence – and schedule a proactive call in the next two weeks. Churn risk caught 90 days before renewal is recoverable; caught 30 days before renewal it rarely is. Third, add a "success milestone" email at the 30-day mark asking whether customers have achieved their primary goal. The responses tell you who is healthy, who is at risk, and what your onboarding sequence is missing.