Customer Success Platform Implementation - Playbooks2 of 3

Customer Success Platform Implementation — Methodology

This document covers the core concepts, frameworks, and calculations behind Customer Success Platform Implementation. It provides the methodological foundation — the "how it works" behind the execution steps.

1) Core Concepts

The Customer Success Platform (CSP) as Operating System

What is it?

A Customer Success Platform is a dedicated software system that aggregates customer data from CRM, product analytics, support, and billing into a single operational hub for Customer Success Managers. Unlike CRM systems (which are sales-oriented record systems) or support tools (which are reactive ticket queues), a CSP is purpose-built for proactive customer management through health monitoring, automated playbooks, and lifecycle orchestration.

Why does it matter?

Without a CSP, CS teams operate reactively - discovering churn risk at renewal time rather than months earlier when intervention could change the outcome. The CSP market reached approximately $4.75 billion in 2024 and is projected to grow at 14-22% CAGR through 2030 [1][2], reflecting the recognition that proactive customer management directly drives Net Revenue Retention.

Key insight:

A CSP is not a reporting tool you log into weekly. It is the daily operating system for every CSM - equivalent to what Salesforce is for an AE. If CSMs are not living inside the platform every day, the implementation has failed regardless of how technically correct it is.

Examples:

Context	Example
Mid-market SaaS ($20M ARR, 15 CSMs)	CSP replaces spreadsheet-based renewal tracking with automated 120-day renewal playbooks, cutting renewal surprises from 30% to under 5%
Enterprise SaaS ($80M ARR, 40 CSMs)	CSP aggregates Salesforce, Pendo, Zendesk, and Stripe data into a single health score, enabling portfolio-level risk management for CS leadership
PLG-hybrid ($10M ARR, 5 CSMs)	CSP identifies product-qualified expansion signals (usage spikes, feature adoption) that trigger CSM outreach for upsell conversations

Common misunderstandings:

Misconception: A CSP replaces the CRM. Reality: The CSP sits alongside the CRM and depends on it as a data source. Salesforce or HubSpot remains the system of record for accounts, contacts, and opportunities. The CSP adds the CS-specific layer: health scores, playbooks, lifecycle tracking, and proactive alerts.
Misconception: Buying a CSP automatically improves retention. Reality: A CSP is infrastructure, not a strategy. Without defined health score methodology, documented playbooks, and trained CSMs who adopt the tool daily, the platform becomes expensive shelfware. Bain & Company found that NRR has declined for 75% of software firms despite increased CS investments [3] - tools without process discipline do not move the needle.

Customer Health Scoring

What is it?

Customer health scoring is a multi-factor model that combines product usage signals, engagement data, support sentiment, and financial indicators into a composite score predicting the likelihood of retention, churn, or expansion for each account. The score typically renders as a traffic-light system (green/yellow/red) with an underlying numeric value.

Why does it matter?

Health scores are the foundation of proactive CS. They determine which accounts get attention, when playbooks trigger, and how leadership assesses portfolio risk. A miscalibrated health score is worse than no score at all - it creates false confidence that leads to missed churn signals and CSM distrust of the platform.

Key insight:

The first health score model is always wrong. Expect to iterate 3-4 times in the first 6 months. The goal of the initial model is to be directionally correct (red accounts should actually be at risk), not perfectly predictive. Start simple with 4-5 signals, validate against known outcomes, then add complexity.

The Framework:

Health scores typically combine four dimensions, each weighted by its predictive power:

Health Score = (Product Usage Weight x Usage Score) +
               (Engagement Weight x Engagement Score) +
               (Support Weight x Support Score) +
               (Financial Weight x Financial Score)

Where all weights sum to 100%

Dimension	What It Measures	Typical Weight Range	Data Sources
Product Usage	Login frequency, feature adoption, license utilization, depth of usage	30-40%	Pendo, Amplitude, Mixpanel, product DB
Engagement	Stakeholder interaction, meeting cadence, executive sponsor access, email responsiveness	20-30%	CRM activity, email tracking, calendar
Support Sentiment	Ticket volume trends, CSAT scores, escalation frequency, resolution time	15-25%	Zendesk, Intercom, Freshdesk
Financial Health	Payment timeliness, contract growth trajectory, renewal proximity, expansion history	10-20%	Billing system (Stripe, Chargebee, Zuora), CRM

Common misunderstandings:

Misconception: More data inputs produce a better health score. Reality: Adding marginal signals introduces noise. Best practice is to monitor metrics without mixing support signals with product adoption signals, measuring leading indicators with short lookup windows [4]. Start with 4-5 high-signal inputs and only add more when the base model is validated.
Misconception: Health scores should be fully automated with no manual override. Reality: CSMs have context that data cannot capture (e.g., champion just left, company was acquired, CEO mentioned competitor in a call). The best CSP implementations allow CSMs to manually adjust scores with documented rationale, which also creates training data for improving the automated model.

Playbook Automation and CTAs

What is it?

Playbooks are pre-defined sequences of tasks triggered by customer signals or lifecycle events. Each playbook generates CTAs (Calls to Action) in the CSM's cockpit - prioritized action items that guide the CSM through a prescribed workflow. CTAs fall into three categories: Risk (churn-related), Opportunity (expansion-related), and Lifecycle (scheduled events like QBRs and renewals) [5].

Why does it matter?

Playbooks are what make CS operations scalable and consistent. Without them, every CSM handles the same scenario differently - one CSM catches a usage drop immediately, another misses it entirely. Playbooks encode institutional knowledge into repeatable workflows and ensure that no critical customer signal goes unaddressed.

Key insight:

Start with 4-5 core playbooks maximum at launch. Every CS team wants 20+ playbooks from day one, but this overwhelms CSMs with noise and makes it impossible to measure which playbooks actually drive outcomes. Launch with: Onboarding, At-Risk Intervention, Renewal Prep, and Expansion Signal. Add more only after proving these four work.

Examples:

CTA Type	Trigger	Playbook	Tasks
Risk	Health score drops below 60 for 2 consecutive weeks	At-Risk Intervention	1) Review account health details 2) Schedule call with CSM manager 3) Prepare save plan 4) Execute executive outreach
Lifecycle	120 days before contract renewal date	Renewal Preparation	1) Review account health and expansion signals 2) Schedule renewal kickoff with champion 3) Prepare renewal proposal 4) Conduct renewal call
Opportunity	License utilization exceeds 85% for 30 days	Expansion Signal	1) Identify additional use cases 2) Schedule expansion conversation 3) Loop in AE for commercial discussion
Lifecycle	Day 1 after contract signed	Onboarding Kickoff	1) Send welcome package 2) Schedule kickoff call 3) Assign onboarding tasks to customer 4) Set 30/60/90 day milestones

Customer Lifecycle Stages

What is it?

Customer lifecycle stages are defined phases a customer moves through from initial purchase to mature, expanding usage. Typical stages include: Onboarding, Adoption, Growth/Value Realization, Renewal, and At-Risk. Each stage has different success criteria, CSM engagement models, and playbook triggers.

Why does it matter?

Lifecycle stages determine what "success" looks like at any given moment and what the CSM should focus on. An onboarding customer needs hands-on guidance and milestone tracking. A mature customer in the growth phase needs strategic value conversations and expansion planning. Treating all customers the same regardless of lifecycle stage leads to misallocated effort.

The Framework:

[Onboarding] --> [Adoption] --> [Growth] --> [Renewal] --> [Onboarding v2]
                    |                           |
               [At-Risk] <--- <--- <--- [At-Risk]
                    |
                 [Churn]

Stage	Duration	Success Criteria	CSM Focus
Onboarding	30-90 days	Key milestones achieved, first value delivered	Hands-on guidance, milestone tracking, stakeholder mapping
Adoption	60-180 days	Feature adoption targets met, daily active usage established	Training completion, workflow integration, expanding user base
Growth	Ongoing	Expanding use cases, increasing license utilization, positive health trend	Strategic reviews (QBRs), expansion conversations, advocate development
Renewal	90-120 days pre-renewal	Renewal decision confirmed, terms negotiated	Value documentation, stakeholder alignment, commercial preparation
At-Risk	Variable	Root cause identified, save plan in progress	Executive engagement, rapid response, recovery playbook execution

2) Decision Frameworks

Platform Selection Matrix

The single most consequential decision in a CSP implementation is platform selection. The wrong platform for the client's maturity level, team size, and technical resources leads to either underutilization (too complex) or outgrown capacity (too simple) within 12 months.

Situation	Recommended Platform	Why
Enterprise, 30+ CSMs, complex account hierarchy, dedicated CS Ops	Gainsight	Most configurable platform for enterprise complexity; advanced ML-based health scoring; deep Salesforce integration. Recognized as a leader in Gartner's inaugural Magic Quadrant for Customer Success Platforms (November 2024) [6]
Mid-market, 10-25 CSMs, need fast time-to-value, moderate complexity	ChurnZero	Faster implementation (6-8 weeks vs 3-6 months for Gainsight); strong analytics at contract and employee levels; good balance of power and usability [7]
Growth-stage, 3-10 CSMs, limited CS Ops resources, need simplicity	Vitally	Fastest implementation (2-4 weeks); intuitive UI with built-in productivity features; strongest fit for teams without dedicated admins [8]
Enterprise with existing Totango investment or HubSpot-centric stack	Totango	Modular design allows phased rollout; stronger HubSpot integration than Gainsight; competitive pricing vs Gainsight for similar enterprise features [9]
Budget-constrained startup, under 5 CSMs, primary CRM is HubSpot	HubSpot Service Hub + custom	Native integration avoids additional platform cost; custom health scoring via calculated properties; limited but sufficient for early CS operations

Scoping Factors

1. CS Team Maturity Level

Early-stage (no formal CS process, spreadsheet tracking) --> Prioritize platforms with guided setup and templates (Vitally, ChurnZero). Keep configuration simple. Focus on getting basic health visibility before automation.
Growth-stage (defined processes, some playbooks, reactive CS) --> Choose platforms with automation depth (ChurnZero, Totango). Invest in CTA automation and lifecycle orchestration.
Mature (formal CS Ops function, data-driven, proactive CS) --> Full enterprise platform (Gainsight). Advanced health scoring with ML, complex playbook trees, executive dashboards.

2. Integration Complexity

Simple (CRM + 1-2 data sources) --> Any platform handles this; weight selection toward usability and price.
Moderate (CRM + product analytics + support + billing) --> Require native integrations for primary sources; evaluate middleware needs (Workato, Tray.io) for secondary.
Complex (CRM + multiple product lines + custom data warehouse + ERP) --> Gainsight or Totango with dedicated implementation partner. Budget 40-60% of annual platform cost for professional services in year one [9].

3. Technical Resources Available

No dedicated CS Ops --> Avoid Gainsight (requires ongoing admin). Choose Vitally or ChurnZero with simpler admin requirements.
Part-time CS Ops --> ChurnZero or Totango, which balance configurability with manageable admin overhead.
Full-time CS Ops + technical resources --> Gainsight or any platform; the team can handle complex configuration and ongoing optimization.

4. Budget Parameters

Under $20K/year --> HubSpot Service Hub or Vitally starter tier. Avoid enterprise platforms that will be underutilized.
$20K-$60K/year --> ChurnZero, Vitally Pro, or Totango. Sweet spot for mid-market implementations.
$60K+/year --> Gainsight, Totango enterprise, or ChurnZero enterprise. Expected range for 30+ CSM teams with complex requirements. Actual average annual costs: Gainsight ~$67K, Totango ~$58K [9].

Health Score Approach Selection

Situation	Recommended Approach	Why
First CSP implementation, limited historical data	Simple Weighted Average (4-5 inputs)	Easy to understand, easy to validate, builds CSM trust before adding complexity
Established CS operation migrating to new platform	Replicate + Enhance existing model	Preserve institutional knowledge, add new data sources available in the CSP
500+ accounts with 2+ years of renewal/churn data	ML-Assisted Scoring (Gainsight)	Enough training data for machine learning to surface non-obvious patterns; use ML as supplement, not replacement
Multiple product lines with different usage patterns	Segment-Specific Models	One health score model cannot fit all products; create separate models per product line with shared engagement and financial dimensions

3) Benchmarks & Standards

How to Use Benchmarks

Benchmarks are guidelines, not rules. Always:

Start with benchmark as baseline
Adjust based on client-specific data (segment, ACV, product type)
Validate against their actual numbers when available
Document deviations and rationale

Retention Metrics Benchmarks

Metric	Low	Typical	High	Notes
Net Revenue Retention (NRR) - Enterprise (ACV >$100K)	<105%	115%	>130%	Enterprise accounts drive NRR through expansion; median NRR for $100M+ ARR companies is 115% [10]
Net Revenue Retention (NRR) - Mid-Market ($25K-$100K ACV)	<100%	108%	>120%	Mid-market is the CSP sweet spot; most implementable segment for health-driven expansion [10]
Net Revenue Retention (NRR) - SMB (<$25K ACV)	<90%	97%	>105%	SMB NRR is typically below 100%; digital-led CS motions are critical here [10]
Gross Revenue Retention (GRR)	<85%	88-90%	>95%	GRR isolates churn from expansion; best-in-class companies achieve 95-100% [11]
Annual Gross Churn Rate	>15%	10-12%	<5%	B2B SaaS average churn dropped to 4.2% in 2024 for well-run operations [12]

Source: SaaS Capital 2024 Retention Benchmarks, ChartMogul SaaS Retention Report, Optifai NRR analysis of 939 companies [10][11][12]

Interpretation:

NRR below 100%: Revenue is shrinking even without logo churn. Urgent problem - the CSP implementation should prioritize churn prevention and contraction reduction before expansion.
NRR above 120%: Expansion is strong - CSP should focus on identifying and scaling the expansion signals that are working.
GRR below 85%: Significant churn problem. CSP health scores need aggressive early-warning configuration (flag at-risk 150+ days before renewal, not 90).

CSP Adoption Benchmarks

Metric	Low	Typical	High	Notes
CSM Daily Active Usage Rate (30 days post-launch)	<50%	65-75%	>80%	Target 80%+ within 30 days; below 50% signals adoption failure requiring intervention
CTA Completion Rate	<40%	55-65%	>80%	Low completion suggests either too many CTAs (noise) or poor workflow fit
Health Score Accuracy (vs actual outcomes)	<60%	70-75%	>85%	Validate quarterly; companies with 80%+ accuracy reduced involuntary churn by 23% [9]
Time to First Value (platform live with core features)	>6 months	8-12 weeks	<6 weeks	Vitally: 2-4 weeks, ChurnZero: 6-8 weeks, Gainsight: 3-6 months [7][8]
Playbook Trigger-to-Action Time	>48 hours	12-24 hours	<4 hours	Time between CTA creation and CSM taking first action; measures workflow integration

Source: Platform vendor benchmarks (Gainsight, ChurnZero, Vitally), Bain & Company CS research [3][7][8]

Quick Reference Thresholds

Question	Good	Warning	Red Flag
Are CSMs using the platform daily?	>80% DAU within 30 days	50-80% DAU	<50% DAU after 30 days
Are health scores predictive?	>80% accuracy vs outcomes	60-80% accuracy	<60% accuracy (worse than random)
Are CTAs being completed on time?	>75% on-time completion	50-75% completion	<50% (CTA fatigue or workflow mismatch)
Is renewal visibility improving?	Renewals flagged 90+ days out	Flagged 30-90 days out	Renewals still surprising the team
Is NRR improving post-implementation?	NRR up 5+ points in 6 months	Stable NRR	NRR declining despite CSP investment

4) Calculations & Scoring

Formula Quick Reference

Calculation	Formula	Example
Composite Health Score	`Sum of (Weight_i x Dimension_Score_i)` for all dimensions	0.35(80) + 0.25(70) + 0.20(90) + 0.20(75) = 78.5
Net Revenue Retention	`(Beginning ARR + Expansion - Contraction - Churn) / Beginning ARR x 100`	($1M + $150K - $30K - $50K) / $1M = 107%
Gross Revenue Retention	`(Beginning ARR - Contraction - Churn) / Beginning ARR x 100`	($1M - $30K - $50K) / $1M = 92%
CSM Capacity Ratio	`Total Managed ARR / Number of CSMs`	$15M / 10 CSMs = $1.5M per CSM
Health Score Accuracy	`Correct Predictions / Total Predictions x 100`	160 correct / 200 accounts reviewed = 80%

Health Score Calculation

Formula:

Health Score = (W_usage x S_usage) + (W_engagement x S_engagement) +
              (W_support x S_support) + (W_financial x S_financial)

Where:
- W = weight (all weights sum to 1.0)
- S = dimension score (0-100 scale)
- Final score range: 0-100

Variables explained:

W_usage (0.30-0.40) = Weight assigned to product usage dimension. Higher for product-led companies.
S_usage (0-100) = Composite of login frequency, feature adoption rate, license utilization. Source: product analytics tool (Pendo, Amplitude, Mixpanel).
W_engagement (0.20-0.30) = Weight for stakeholder engagement. Higher for enterprise/high-touch models.
S_engagement (0-100) = Composite of meeting cadence, email responsiveness, executive sponsor access, NPS response. Source: CRM activity logs, calendar data.
W_support (0.15-0.25) = Weight for support sentiment. Higher for complex products with frequent support needs.
S_support (0-100) = Composite of ticket volume trend (declining = good), CSAT scores, escalation frequency. Source: Zendesk, Intercom.
W_financial (0.10-0.20) = Weight for financial health. Higher for companies with payment risk or contraction patterns.
S_financial (0-100) = Composite of payment timeliness, contract growth trajectory, expansion history. Source: billing system (Stripe, Chargebee, Zuora).

Worked Example:

Scenario: Mid-market B2B SaaS account, $50K ARR, 18 months into contract

Given:
- Product Usage Score = 75 (good login frequency, moderate feature adoption, 70% license utilization)
- Engagement Score = 60 (missed last QBR, champion responsive but no exec sponsor contact in 3 months)
- Support Score = 85 (low ticket volume, high CSAT, no escalations)
- Financial Score = 90 (pays on time, expanded 10% last quarter)

Weights (mid-market standard):
- Usage: 0.35
- Engagement: 0.25
- Support: 0.20
- Financial: 0.20

Calculate:
- (0.35 x 75) + (0.25 x 60) + (0.20 x 85) + (0.20 x 90)
- = 26.25 + 15.0 + 17.0 + 18.0
- Health Score = 76.25 (Yellow - monitor closely)

Validation:

Score should be 0-100. If outside this range, check weight normalization.
Green: 80-100 (healthy, maintain engagement)
Yellow: 50-79 (at-risk, increase touchpoints, investigate low-scoring dimensions)
Red: 0-49 (critical, trigger intervention playbook immediately)
If a "known healthy" account scores below 60, the model needs recalibration.
If a "recently churned" account had a score above 70 before churn, the model is missing key signals.

Health Score Threshold Calibration

Tier Thresholds (starting point - adjust per client):

Tier	Score Range	Action
Green (Healthy)	80-100	Standard engagement cadence; expansion opportunities
Yellow (At-Risk)	50-79	Increased monitoring; schedule proactive check-in; investigate declining dimensions
Red (Critical)	0-49	Immediate intervention; escalate to CS leadership; execute save playbook

Calibration Process:

Apply initial thresholds to entire customer base
Distribution should be approximately: 60-70% Green, 20-30% Yellow, 5-15% Red
If distribution is heavily skewed (e.g., 90% Green), thresholds are too generous - raise them
Validate against the past 12 months of churn: did red accounts actually churn? Did churned accounts show red before leaving?
Adjust weights and thresholds based on validation, repeat quarterly

CSP ROI Calculation

Formula:

Annual CSP ROI = [(Churn Reduction Value + Expansion Revenue Increase + CSM Efficiency Gains)
                  - (Platform Cost + Implementation Cost)]
                 / Total Investment x 100

Worked Example:

Scenario: 500-account SaaS company, $30M ARR, implementing ChurnZero

Given:
- Current annual churn: 12% ($3.6M lost)
- Expected churn reduction with CSP: 25% reduction [9]
- New churn: 9% ($2.7M lost)
- Churn Reduction Value: $900K
- Expansion revenue increase (from better signal detection): $300K
- CSM efficiency gains (time saved on manual tasks): 5 hrs/CSM/week x 15 CSMs x $50/hr x 50 weeks = $187.5K
- Platform cost: $45K/year
- Implementation cost (year 1 only): $25K

Year 1 ROI = ($900K + $300K + $187.5K - $45K - $25K) / $70K = 1,882%
Year 2+ ROI = ($900K + $300K + $187.5K - $45K) / $45K = 2,983%

Validation:

Well-implemented CSPs deliver 150-400% ROI [9]. If the calculation shows less than 100%, double-check churn reduction assumptions or consider whether the company has enough ARR to justify the investment.
The 25-32% churn reduction figure comes from industry data on properly deployed platforms [9]. Adjust down for less mature CS operations.

5) Edge Cases & Deep Dives

Edge Case 1: Multi-Product Company with Different Usage Patterns

Scenario:

Client sells 3 distinct products (e.g., a core platform, an analytics add-on, and a professional services portal). Each product has different usage patterns, different definitions of "active user," and different health indicators. Some customers use all three, some use only one.

Challenge:

A single health score model cannot accurately represent health across products with fundamentally different usage patterns. A customer might be highly engaged with the core product but never logging into the analytics add-on they purchased.

Approach:

Create a separate health score model per product, each with product-specific usage metrics and weights
Create a composite "Account Health Score" that aggregates product-level scores weighted by each product's ARR contribution
Configure CTAs at both levels: product-level (low adoption of specific product) and account-level (overall health decline)

Key validation:

Review 10 multi-product accounts manually. Does the composite score align with the CSM's gut assessment? If not, adjust product weights in the composite formula.

Edge Case 2: PLG-Hybrid Company (Product-Led + Sales-Assisted)

Scenario:

Client has a freemium or free-trial funnel generating self-serve users alongside an enterprise sales motion. The CSP needs to handle both product-qualified accounts (converted from PLG) and sales-qualified accounts (enterprise deals). These two cohorts have fundamentally different onboarding experiences, engagement patterns, and expansion paths.

Challenge:

PLG accounts may show high product usage but low stakeholder engagement (no champion identified, no executive sponsor). Enterprise accounts may show strong engagement signals but lower product usage (IT-managed deployment, not individual user adoption). A single health model misrepresents both.

Approach:

Segment accounts by acquisition channel (PLG vs Sales-Led) as a first-class dimension in the CSP
Create separate health score models per segment with different weight distributions:
- PLG: Usage 50%, Engagement 15%, Support 20%, Financial 15%
- Enterprise: Usage 25%, Engagement 35%, Support 15%, Financial 25%
Create PLG-specific playbooks: "Champion Identification" (triggered when PLG account crosses ARR threshold), "Stakeholder Mapping" (no executive contact after 60 days)
Set different lifecycle stage durations per segment (PLG onboarding may be self-serve and shorter)

Key validation:

Compare churn rates between PLG and sales-led cohorts. If the PLG model flags different accounts than the enterprise model would, the segmentation is working correctly.

Edge Case 3: Data Quality Issues Discovered Mid-Implementation

Scenario:

During CRM integration setup, the team discovers that 40% of accounts have missing or stale data: incorrect renewal dates, no ARR field populated, contacts with wrong roles, or duplicate accounts. The CRM was never cleaned before CSP implementation was initiated.

Challenge:

Health scores calculated on bad data produce bad results. Launching with inaccurate health scores destroys CSM trust in the platform - the single biggest adoption risk. But waiting for a full CRM cleanup could delay the project by months.

Approach:

Triage data quality into three buckets:
- Critical (blocks health scoring): ARR values, renewal dates, primary contact. These must be fixed before launch.
- Important (degrades accuracy): Account hierarchy, contact roles, product ownership. Fix during pilot phase.
- Nice-to-have (enhances reporting): Industry classification, territory assignments. Fix post-launch.
Run a focused data cleanup sprint on Critical fields only - typically 2-3 weeks with CS Ops + CRM admin
Configure health score to gracefully handle missing data: if a dimension's data is unavailable, exclude that dimension and redistribute its weight to available dimensions (document this in the scoring methodology)
Flag accounts with data quality issues in the CSP using a custom field ("Data Confidence: High/Medium/Low")
Create a CTA playbook: "Data Quality Review" triggered for accounts with Low data confidence, assigning the CSM to verify and update key fields during their next customer interaction

Key validation:

After the focused cleanup sprint, re-run health scores and compare distribution to expected ratios (60-70% Green, 20-30% Yellow, 5-15% Red). If the distribution is plausible and known at-risk accounts score Red, proceed to pilot.

Edge Case 4: Enterprise Client with Restricted Data Access

Scenario:

Client is in a regulated industry (financial services, healthcare) or has strict InfoSec policies that prevent product usage data from being sent to a third-party CSP. API access requires security review, and some data sources cannot be integrated at all.

Challenge:

Without product usage data (typically the highest-weighted health score dimension), the health model loses its most predictive signal. The CSP implementation scope shrinks significantly, and the business case weakens.

Approach:

Build health score model around available dimensions only: engagement, support, and financial signals
Increase engagement weight to 40-45% since it becomes the primary leading indicator
Use CSM-reported product adoption assessments as a manual proxy for usage data (structured form completed during QBRs)
Explore on-premise or VPC-deployed CSP options if the client requires it (Gainsight offers private cloud deployment)
Investigate whether aggregated/anonymized usage metrics (not raw user-level data) can pass security review - often aggregates are approved when individual records are not

Key validation:

Compare health score predictions from the reduced model against actual renewal outcomes for 2 quarters. If accuracy is below 60%, the model is not adding value and needs supplemental data sources or a different approach.

Edge Case 5: Migrating from One CSP to Another

Scenario:

Client already has a CSP (e.g., Totango) but is migrating to a new one (e.g., Gainsight) due to outgrowing the current platform, organizational mandate, or dissatisfaction with current tooling.

Challenge:

Migration risks losing historical health score data, playbook configurations, CTA history, and CSM familiarity. The "dip" in productivity during migration can last months and may cause real churn during the transition period.

Approach:

Export all possible historical data from the current platform: health score snapshots, CTA completion records, playbook definitions, account notes
Run both platforms in parallel for 4-6 weeks during migration (old platform read-only, new platform as primary)
Replicate existing health score model first before enhancing it - CSMs need familiarity before change
Migrate playbook logic (not just structure) including trigger conditions, task sequences, and escalation paths
Prioritize data continuity over feature parity - historical context is more valuable than new features in the short term

Key validation:

After parallel run period, compare health score distributions between old and new platforms. Scores should correlate (not necessarily match) for the same accounts. If correlation is weak, the new model needs recalibration before decommissioning the old platform.