SaaS Metrics — MRR, ARR, Churn, LTV, CAC and North Star for Startups That Generate Revenue

Your SaaS startup is growing, but you don't know which levers to pull. The bank balance goes up and down. Investors ask for numbers, but your reports mix invoices and fixed costs. You feel like a shopkeeper who only sees the cash register, not the margin. We at Meteora Web come from accounting — balance sheets, double-entry bookkeeping, VAT. That's why when we talk about SaaS metrics, we don't stop at marketing. We look at real numbers. The ones that separate a sustainable business from a bubble.

This guide gets straight to the point: what MRR, ARR, churn, LTV, CAC and North Star metric actually mean, how to calculate them without a math degree, and most importantly, what to do with those numbers today to grow your startup.

Why are SaaS metrics different from a physical store or e-commerce?

In a clothing store — and we managed the ERP of one — accounting is linear: buy goods, sell them, calculate margin by season. In SaaS, you sell a subscription. The customer pays today, but you must keep them for months or years. The initial revenue covers only part of it. The real profit comes if the customer stays. That's where the metric you can't ignore comes in: churn.

If you lose 5% of customers every month, your business is a leaky bucket. You can fill it as much as you want, but the water keeps dropping. That's why SaaS metrics need to be monitored in real time, not once a quarter.

Actions you can take now

• Export a list of active customers with first payment date and last payment date.
• Calculate monthly churn rate: customers lost in month / customers at start of month × 100.
• If you are above 5% monthly, you have a structural problem: stop new acquisitions until you fix retention.

How to calculate MRR and ARR — and why they are not the same thing?

MRR (Monthly Recurring Revenue) is your daily bread. It's the recurring revenue of all active customers, normalized to a monthly basis. If you have yearly plans, divide the annual value by 12. For quarterly plans, divide by 3. ARR (Annual Recurring Revenue) is simply MRR × 12. It sounds simple, but the most common mistake is including one-time fees (setup, consulting) or non-recurring upgrades. We've seen this in dozens of startup financials: they inflate ARR, fooling founders and investors.

Calculate it with a Python script (runs on any server or notebook)

def calculate_mrr(subscriptions):
    """
    subscriptions: list of dict with 'plan_type' ('monthly','yearly','quarterly') and 'amount'
    """
    total = 0
    for sub in subscriptions:
        if sub['plan_type'] == 'monthly':
            total += sub['amount']
        elif sub['plan_type'] == 'yearly':
            total += sub['amount'] / 12
        elif sub['plan_type'] == 'quarterly':
            total += sub['amount'] / 3
    return round(total, 2)

# Example
subscriptions = [
    {'plan_type': 'monthly', 'amount': 29},
    {'plan_type': 'yearly', 'amount': 299},
    {'plan_type': 'quarterly', 'amount': 79},
]
print(f"MRR: ${calculate_mrr(subscriptions)}")  # Output: MRR: $84.42

Use this script on a JSON file from your billing system. Run it weekly.

What is the real customer acquisition cost (CAC) for SaaS and how to avoid mistakes?

CAC = total sales and marketing costs in a period / number of new customers acquired in that period. Costs must include salaries of sales and marketing people, ad spend, CRM subscriptions, ad platform costs. Typical mistake: forgetting the founder's time spent on demos. At Meteora Web, when we work with startups, we ask for man-hours too — because the founder's time is not free. A high CAC is not a problem if the customer stays long. But if CAC exceeds the first year of revenue, you are losing money on every customer.

Calculate LTV/CAC ratio right now

• Gather data from the last 3 months: sum all sales and marketing costs (including your salary).
• Divide by new customers acquired. You get CAC.
• Calculate LTV (see next paragraph).
• If LTV/CAC < 3, your acquisition channels are too expensive or your product doesn't retain enough.

How to calculate LTV (Lifetime Value) without guessing?

The basic formula: LTV = average MRR per customer / monthly churn rate × gross margin %. Gross margin in SaaS is usually high (80-90%) but must be calculated: revenue – infrastructure costs (servers, third-party APIs, direct support). Example: avg MRR per customer = $50, monthly churn = 3%, margin 85%. LTV = (50 / 0.03) × 0.85 = $1,416.67. That's how much you can afford to spend to acquire one customer. We always start with one question: how much does it cost and how much does it return? The rest comes after.

Warning: churn is not linear. First months have higher drop-off. Calculate churn by cohort (customers who entered in the same month) and use that for a more realistic LTV.

Python script for cohort LTV

def ltv_by_cohort(cohorts, margin=0.85):
    """
    cohorts: dict with 'month' as key, list of monthly average revenue per active customer
    """
    for month, revenues in sorted(cohorts.items()):
        mrr_avg = sum(revenues) / len(revenues) if revenues else 0
        churn_rate = 0.03  # estimate, get from your data
        ltv = (mrr_avg / churn_rate) * margin
        print(f"Cohort {month}: avg LTV = ${round(ltv, 2)}")

North Star Metric — the one metric that drives real growth

MRR and churn tell you where you are. The North Star Metric tells you where you're going. It's the metric that, when improved, naturally grows everything else. For Spotify it's "listening time". For Slack it's "messages sent". For an Italian SaaS PMI it might be "active usage sessions per customer" or "feature X used at least once a day". How to find it? Ask yourself: what action of my customers is correlated with retention and referral? Then make it measurable. We at Meteora Web built a proprietary platform to manage social presence for multiple clients: our North Star was "scheduled posts published without errors". It sounds trivial, but when that metric went up, customers stayed and invited others.

Define yours today

• Pick 10 customers who have renewed for more than 12 months.
• Analyze what common action they do more than others (e.g., number of logins, hours of usage, reports generated).
• That action is your North Star candidate. Add it to your main dashboard.

What to do next with these SaaS metrics

You don't need an enterprise ERP. You need a well-made spreadsheet or the Python scripts I gave you. Concrete actions for this week:

1. Calculate MRR and ARR with the script above: if your declared ARR includes one-time fees, correct your investor deck immediately.
2. Calculate CAC including your own time. If LTV/CAC is below 3, cut ad campaigns and invest in content or referrals.
3. Identify your North Star metric using the 10 loyal customers method.
4. Automate monthly churn calculation: run a cron job on the first of every month (we use a PHP script with SQL queries).
5. Schedule a call with an experienced SaaS advisor — our pillar guide on startup and product management has the full methodology.

A website measures in revenue, not compliments. SaaS metrics are no exception.