SaaS Metrics Health Check: MRR, Churn, CAC and LTV

SaaS Metrics Health Check is an AI skill that takes your raw business numbers — recurring revenue, customer counts, churn, and sales and marketing spend — calculates the metrics that matter, labels each one healthy, watch, or critical against a benchmark, and tells you what to fix first. It's direct by design: if a metric is bad, it says so.

What it does

It's adapted from a skill called SaaS Metrics Coach, which is written to act like a senior SaaS finance adviser, and it works in four steps.

1. It collects your inputs in one go. Revenue (this month's and last month's monthly recurring revenue, expansion, and churned revenue), customers (active, new, and lost this month), and costs (sales and marketing spend, gross margin). It works with partial data and says plainly what's missing and what it's assuming.

2. It calculates the core metrics. Annual recurring revenue, month-over-month growth, monthly churn rate, customer acquisition cost, lifetime value, the LTV:CAC ratio, CAC payback period, and net revenue retention.

3. It benchmarks each one — for your kind of company. Every metric gets a value, a benchmark range, and a status label. The benchmark depends on your segment (enterprise, mid-market, SMB, or product-led) and stage (early, growth, or scale), because the same churn figure can be alarming for an enterprise product and ordinary for a self-serve one. If your market isn't clear, it asks before scoring.

4. It prioritises. It picks the two or three worst metrics — never more, because a longer list stalls action — and for each gives what's happening in one sentence, why it matters, and a few specific actions for this month.

The output is always the same shape: a metrics table, a short overall picture, the priority issues, what's genuinely working, and a single metric to move over the next 90 days with a numeric target.

A worked example

This one comes from the original skill's own examples. Say monthly recurring revenue is 22,000, down from 23,500 last month. You have 80 customers, lost 9 this month and gained 6, spent 15,000 on ads, and run a 65% gross margin.

• Growth: 22,000 against 23,500 is about −6.4% month over month.
• Churn: 9 lost out of 80 is 11.25% in a single month.
• Lifetime value: revenue per account is 275, or about 179 of gross profit at a 65% margin. Divided by that churn rate, a customer is worth roughly 1,590 over their lifetime.
• Acquisition cost: 15,000 spent for 6 new customers is 2,500 each.
• LTV:CAC: about 0.64 — every new customer costs more to win than they'll ever return.

The skill marks all three as critical and puts churn first: there's no point spending more to acquire customers who leave this fast. That ordering is the useful part. A spreadsheet shows the same numbers; it doesn't tell you which fire to put out first.

What it won't do

• connect to your billing, payments, or accounting systems — it uses only the numbers you give it
• make transactions or change anything in your business
• give personalised financial or investment advice

Its benchmark ranges are general reference points, not guarantees, and the output is informational. For decisions with real financial consequences, bring the report to someone who knows your business.

When to reach for it

• You're preparing a monthly investor or board update
• You're not sure whether your churn is normal for your market
• You're about to increase acquisition spend and want to check the economics first
• You changed pricing and want to see the effect in the numbers
• You have the figures in a spreadsheet but no clear read on what they mean

What's actually inside

One skill. No connectors, no account access, no API keys.

The original folder carries more: Python calculators for the core metrics, a quick-ratio calculator for growth efficiency, a 12-month unit-economics simulator for "what if" scenarios, reference files with formulas and benchmark ranges, and a blank input template. The Taku listing describes a skill that calculates from the data you supply, with benchmark labels, stated assumptions, and warnings about missing inputs.

Setting it up

The GitHub route. The original is in finance/skills/saas-metrics-coach inside claude-skills, a public collection of several hundred skills for Claude Code, Codex, Gemini CLI, and other agents. Copy the whole folder — scripts and references included — into your AI tool's skills directory and restart it. The calculators need Python; without them, the skill falls back to its formula reference.

The Taku route. Open it in Taku and give it your numbers.

Who made it

The original SaaS Metrics Coach credits Abbas Mir as its author and lives in Alireza Rezvani's MIT-licensed claude-skills collection. SaaS Metrics Health Check was published to the Taku marketplace by Alireza Rezvani.

You don't need to be technical to use this

You don't need to understand GitHub, Codex, or Claude Code to use this skill. Those are the usual homes for AI skills, and they assume you're comfortable with repositories and Python — which is a lot to ask of a founder who just wants to know whether their churn is a problem.

Taku is a desktop app built for exactly that: find something useful someone already made, open it in your own workspace, and use it with no setup. Browse the free app library for more, or read AI tools for startups for what's worth adopting at each stage. When the report needs to make sense to people outside finance, the Plain Language skill rewrites it without losing the numbers.

Download Taku and run this month's numbers through it. Taku is in Beta, and the Mac app is available now; there's also an experimental, unsigned Windows build for Windows 10 and 11.

FAQ

Which metrics does it calculate?

Annual recurring revenue, month-over-month revenue growth, monthly churn rate, customer acquisition cost, lifetime value, the LTV:CAC ratio, CAC payback period, and net revenue retention.

Do I need complete data?

No. It works with partial data, states what's missing and what it assumed, and asks for the single most useful missing figure.

Does it connect to my billing or accounting system?

No. It uses only the numbers you provide and doesn't access any accounts.

Is this financial advice?

No. The output is informational, with general benchmark ranges. It isn't personalised financial or investment advice.

Is it free?

The skill is free. You need an AI tool that can run it. For what Taku costs, see pricing.

Key points

• Turns raw SaaS numbers into growth, churn, CAC, LTV, payback, and NRR.
• Labels each metric healthy, watch, or critical against segment- and stage-specific benchmarks.
• Caps priority issues at three, each with concrete actions for this month.
• Works with partial data and says what's missing.
• Uses only numbers you supply — no system access, not financial advice.