Gainsight, ChurnZero, and the limits of the health score

Researched and published in December 2026. Vendor capabilities change — verify current details directly.

Gainsight, ChurnZero, and the limits of the health score

A health score is a prediction, and a good one is genuinely useful. It orders a CSM's queue, it triggers a play, it gives a leader a portfolio view. If you run a customer success team of any size without one, you're allocating attention by whoever emailed most recently.

But a health score answers exactly one question — who is at risk — and teams routinely ask it a second question it cannot answer: why, and what do we do about it.

That gap is not a product flaw in Gainsight or ChurnZero. It's a category boundary. This post is about where the boundary sits, what the platforms do well inside it, and what to do about the part that falls outside.

Key takeaways

• Health scores are predictive, not diagnostic. They rank risk; they don't establish cause.
• A composite score is a weighted average of proxies, and the weights are usually a guess that hardened into a policy.
• Most causes of B2B churn originate outside CS — in product, billing, or support — which is why a CS-owned score can identify risk it cannot resolve.
• Gainsight, ChurnZero, Totango, Catalyst, Planhat and Vitally are genuinely differentiated by company size, implementation weight, and price. Pick on those.
• The diagnostic layer is a different product, and it usually runs alongside the CS platform rather than instead of it.

What a health score actually is

Almost always a weighted composite: product usage, support ticket volume, NPS or survey responses, executive engagement, invoice status, sometimes contract data.

Two structural properties follow, and both matter.

The weights are chosen, not derived. Someone decided usage is 40% and support volume is 20%. Occasionally that's fitted to historical churn; more often it was a workshop in year one and nobody has revisited it. Either way, the score is partly a statement about your team's beliefs.

Composites lose information on the way in. Two accounts scoring 62 can be entirely different situations — one with strong usage and terrible support experience, one with light usage and a delighted champion. The score treats them as equivalent, and the CSM has to open the record and reconstruct what's actually going on.

Causal finding this is arithmetic, not observation. Any weighted composite maps many input states onto one output value. That's what compositing is for, and it means the score cannot carry the information needed to determine cause.

The three questions a score can't answer

Why is this account at risk? The score dropped because usage fell. Usage fell because the champion left, or because a release broke their main workflow, or because they hired a team that prefers a competitor, or because their own business is shrinking. Four completely different situations, one identical score movement, four different correct responses.

Who has to fix it? If the cause is a product regression, the CSM cannot fix it. They can escalate, but escalation is a social process — finding the right person, being persuasive, competing with the roadmap. The score generated a task for the one team least able to resolve most of its causes.

Did our intervention work? The account renewed. Was that the QBR, or would they have renewed anyway? Without a comparison group, the CS team's attributed saves include every account that was never really at risk — and health-score-triggered plays are systematically aimed at accounts selected for having a signal, which is the classic selection problem.

The platforms, honestly

Gainsight

Best for: enterprise CS organizations with dedicated CS Ops.

The most complete platform in the category. Deep configurability, mature playbook engine, strong analytics, extensive integrations, and the largest practitioner community and body of methodology in customer success.

Where it wins: breadth, enterprise readiness, and the sheer amount of institutional knowledge in the ecosystem. Its published research is genuinely a public good for the profession. Where it stops: implementation is heavy and usually needs dedicated ops headcount. Cost puts it out of range for many mid-market teams. Right for you if: you have a large CS org, complex segmentation, and someone whose actual job is administering it.

ChurnZero

Best for: mid-market SaaS with a fast implementation requirement.

Focused on the CSM's daily workflow rather than on executive analytics. In-app messaging is a genuine strength — the ability to communicate inside the product where usage is actually happening.

Where it wins: faster to value, strong in-app engagement, good fit for high-volume CSM books. Where it stops: less depth in enterprise-grade reporting and complex hierarchies. Right for you if: you're mid-market, want CSMs productive in weeks, and don't need heavy configurability.

Totango

Best for: teams that want to start narrow and expand.

Modular, with pre-built programs you can adopt individually rather than implementing a whole platform at once.

Where it wins: low-commitment entry, decent free tier, quick to pilot. Where it stops: the modular approach can fragment as you scale. Right for you if: you're building a CS function and want to avoid a large upfront commitment.

Catalyst, Planhat, Vitally

Strong options with distinct centres of gravity: Catalyst for sales-led organizations with tight CRM coupling, Planhat for data-model flexibility and non-standard business models, Vitally for product-led companies where usage data is the primary signal and design quality matters to adoption.

Comparison

The bottom three rows are the category boundary, not a criticism. These are customer success platforms and they're built for a CS team.

Where the diagnostic layer sits

Here's the situation that recurs.

An account's health score drops. The CSM investigates and finds usage fell in one workflow. They escalate to product. Product finds a release changed that workflow. By then it's been six weeks, the renewal is in four, and the same release affected two hundred other accounts whose scores also dropped — each investigated separately by a different CSM, arriving at the same conclusion independently.

Strong association in the businesses we work with, when a health-score decline is traced to its cause, that cause frequently originates outside CS and affects a cohort rather than an account. Two hundred individual investigations is the observable symptom of a missing cohort-level diagnostic layer.

That's what Flolyt does. Not health scores — cause, cohort, dollar figure, and the cross-functional Room where the team that can actually fix it is present.

Is Flolyt right for you? Probably, if:

• Your CS team escalates the same underlying cause repeatedly
• Churn causes originate in product, billing, or support more often than in CS
• You can't distinguish saves from accounts that were never at risk
• The same retention problem gets rediscovered annually

Probably not, if you need CSM workflow tooling, task management, QBR generation, or in-app engagement. Flolyt doesn't do those, and the platforms above do them well.

Which should you choose?

• Enterprise CS org with dedicated ops → Gainsight
• Mid-market, need CSMs productive fast → ChurnZero
• Building CS from scratch, low commitment → Totango
• Product-led, usage is the primary signal → Vitally
• Sales-led with heavy CRM coupling → Catalyst
• Unusual data model → Planhat
• You know who's at risk and not why → that's the diagnostic gap, and it's a different tool
• Your churn originates outside CS → keep your CS platform, add a layer above it

Frequently asked questions

Is Flolyt a Gainsight competitor? No. Gainsight manages the CS function — CSM workflows, playbooks, QBRs, portfolio views. Flolyt diagnoses why revenue is leaking across all ten lifecycle stages and coordinates the cross-functional fix. Customers run both.

Can't I just build better health scores? Better scores improve ranking, which is worth doing. They don't produce causes, because a composite by construction cannot carry the information needed to determine one.

Why do you recommend competitors in your own comparison? Because these are good products solving a real problem we don't solve. Telling you Vitally is a strong fit for product-led CS costs us nothing — we're not competing for that budget.

What's the overlap in practice? Some. Both surface at-risk accounts. The difference is what happens next: a CS platform routes it to a CSM with a playbook; Flolyt establishes cause, quantifies it, and opens a Room with the team that owns the source stage.

Do I need both? If your CS team is functioning and your problem is that churn causes live outside CS, then yes. If your CS team lacks basic workflow tooling, start there — that's a more urgent gap.