Stop Guessing on Subscription

Why Subscription Metrics Matter More Than They Appear To

Subscription revenue compounds. A 2% monthly churn rate looks acceptable until you realize it means losing 22% of cohort value in a year. A 5% month-over-month growth rate in MRR can mask a collapsing retention curve if acquisition is inflated by one-time discounts.

For Shopify DTC brands, subscription is often the margin engine. One-time purchases carry CAC friction and payment processing costs. Subscriptions reduce both. But subscription also creates obligation - the brand must deliver every cycle, or churn accelerates.

The mistake: treating subscription as a product feature rather than a business model. Operators need to separate subscription acquisition (how many new subscribers, at what price, via which channel) from subscription retention (how many stay, at what cost to service, with what expansion opportunity).

Core Subscription Metrics and Thresholds

Monthly Recurring Revenue (MRR) is the foundation. Calculate it as: (Number of Active Subscribers) × (Average Revenue Per Subscriber Per Month). Track this by cohort, channel, and product. A healthy DTC brand should see MRR grow 3 - 7% month-over-month in steady state, with new cohorts contributing 15 - 25% of total MRR in their first year.

Churn Rate (Monthly): (Subscribers Lost in Month / Subscribers at Start of Month) × 100. Threshold: below 5% is strong, 5 - 10% is acceptable, above 10% signals product or pricing friction. Track by cohort age - new subscribers churn 2 - 3x faster than 12+ month cohorts.
Net Revenue Retention (NRR): (MRR End of Period - Churn + Expansion) / MRR Start of Period. Expansion includes upsells, plan upgrades, and add-ons. NRR above 100% means existing customers generate more revenue over time. Threshold: 95%+ is healthy, below 90% means churn outpaces growth.
Customer Acquisition Cost for Subscription (CAC-Sub): Total marketing spend attributed to subscription acquisition / number of new subscribers acquired. Compare to Lifetime Value (LTV). Threshold: LTV / CAC-Sub should be 3:1 or higher. If it's 2:1, subscription acquisition is too expensive.
Payback Period: Months to recover CAC from subscription gross margin. Formula: CAC-Sub / (Average Monthly Gross Margin per Subscriber). Threshold: 12 months or less. Above 18 months, the business model breaks - you're funding customer acquisition with debt.
Expansion Rate: (Revenue from existing subscribers in Month N that didn't exist in Month N-1) / MRR at start of Month N. Threshold: 2 - 5% is healthy. Below 1% means no upsell motion; above 10% suggests pricing power or strong product-market fit.

Failure Modes: What Kills Subscription Cohorts

Cohort collapse happens when a single cohort's churn rate exceeds 15% in months 2 - 4. This signals product-market fit failure, not acquisition failure. The fix is not more marketing - it's understanding why that cohort doesn't stick. Common causes: wrong audience (paid ads attracted price-hunters, not repeat buyers), wrong product (subscription doesn't match use case), wrong onboarding (customers don't understand value before first charge).

Payback period creep occurs when CAC-Sub rises faster than gross margin. This happens when: (1) paid channels saturate and brands shift to higher-CAC channels, (2) discounts increase to hit growth targets, (3) product cost rises without price increase. If payback period exceeds 18 months, the subscription model is funding acquisition with future cash flow - unsustainable.

Expansion collapse happens when NRR drops below 95%. This usually means: (1) no upsell motion (product doesn't offer higher tiers or add-ons), (2) pricing is wrong (customers feel they're overpaying), (3) product quality degraded (customers downgrade or churn instead of expand). Fix: audit the top 20% of customers by MRR. Why aren't they expanding? What would make them upgrade?

Seasonal churn spikes occur in Q1 (New Year's resolutions fail, budget resets) and post-holiday (gift subscriptions expire). Plan for 2 - 3x normal churn in these windows. Automate win-back campaigns 30 days before expiry; don't wait for churn to happen.

What to Automate and What to Keep Human

Automate the mechanical work: churn prediction, dunning (retry failed payments), win-back email sequences, and cohort reporting. These are rules-based and repeatable. A customer with 3+ failed payment attempts should enter a dunning sequence automatically. A subscriber who hasn't logged in for 60 days should receive a re-engagement email. A cohort that hits 12% churn in month 3 should trigger an alert to the product team.

Keep human: cohort analysis, pricing decisions, and product changes. Why did this cohort churn 18% instead of 8%? Was it audience, product, or timing? A human needs to investigate. Should we raise prices? That's a business decision, not an algorithm. Should we add a new subscription tier? That requires understanding customer segments, not just running a test.

The hybrid zone: retention campaigns. Automate the trigger (customer at risk of churn based on usage, engagement, or payment history). Keep human: the message. A generic "Come back!" email underperforms. A message that says "We noticed you haven't used X feature - here's how to get the most value" performs 3 - 5x better. Template the structure, personalize the insight.

Connecting Subscription Metrics to CAC and Cohort Health

Subscription CAC must be evaluated against [cohort retention curves](/blog/stop-guessing-on-cohorts). A channel that acquires subscribers at $15 CAC looks great until month 4, when that cohort's churn rate hits 18%. Meanwhile, a $25 CAC channel produces a cohort with 6% monthly churn. The second channel is cheaper on a lifetime basis, even though CAC is higher.

Link [CAC](/blog/stop-guessing-on-cac) back to subscription payback period. If CAC-Sub is $100 and gross margin per subscriber is $8 / month, payback is 12.5 months. If that cohort's churn rate is 8% monthly, only 44% of the cohort survives to payback. The true CAC is actually $227 ($100 / 0.44). This is why cohort analysis matters - it forces you to account for churn in your unit economics.

Track subscription acquisition by channel and product. Paid search might acquire subscribers at $12 CAC with 7% monthly churn. Organic might acquire at $0 CAC with 4% monthly churn. Organic looks better, but if organic volume is capped at 20 subscribers / month and paid can scale to 200, the business needs both. The decision is: at what volume does paid CAC exceed acceptable payback?

Reporting and Decision Rules for Subscription Operations

Build a weekly subscription dashboard with five metrics: (1) MRR (current month), (2) New Subscribers (weekly count), (3) Churn Rate (7-day rolling average), (4) Payback Period (by acquisition channel), (5) NRR (trailing 3-month). These five numbers tell you if the business is healthy or broken.

Set decision rules: If churn rate exceeds 12% for two consecutive weeks, pause all paid acquisition and investigate. If payback period exceeds 15 months, reduce CAC spend by 20% and audit pricing. If NRR drops below 92%, freeze new feature work and focus on retention and expansion. These aren't guesses - they're thresholds based on unit economics.

Cohort retention table: rows are cohort start dates (monthly), columns are months since acquisition (0, 1, 2, 3, 6, 12). Each cell shows the percentage of that cohort still active. This table reveals patterns - which channels produce sticky cohorts, which months produce weak cohorts, which product changes affected retention.
Channel attribution for subscription: tag every new subscriber with acquisition channel, campaign, and creative. At month 3, month 6, and month 12, calculate churn rate and NRR by channel. Reallocate budget toward channels that produce the highest LTV, not the lowest CAC.
Expansion audit: monthly, identify the top 20% of subscribers by MRR. Pull their usage data, plan tier, and expansion history. Why haven't they expanded? Are they maxed out (already on highest tier)? Are they underutilizing (could benefit from upgrade but don't know it)? Are they price-sensitive (would churn if you ask)? This informs product and pricing strategy.

Subscription Pricing and [Attribution](/blog/stop-guessing-on-attribution) Complexity

Subscription pricing is harder to optimize than one-time pricing because changes affect cohort economics retroactively. If you raise prices mid-cohort, you increase MRR but may accelerate churn. If you lower prices, you reduce churn but compress margins. The decision requires understanding price elasticity by customer segment.

Attribution becomes critical when subscription includes discounts, trials, or bundled offers. A customer who starts on a $9.99 trial and upgrades to $29.99 after 7 days should be attributed to the trial offer, not the full price. But if that customer churns at month 3, was the trial offer a bad acquisition channel (attracted low-intent customers) or a good one (converted high-intent customers who otherwise wouldn't have tried)? The answer depends on the cohort's full lifetime, not just the first month.

Test pricing changes on new cohorts, not existing ones. Run a 50 / 50 split: 50% of new subscribers see price A, 50% see price B. Track both cohorts for 6 months. Measure CAC, churn, and LTV. The cohort with higher LTV wins, even if CAC is higher. This is how you find the price that maximizes unit economics, not just revenue.

Questions

FAQ

What's the difference between churn rate and NRR?

Churn rate measures the percentage of customers lost in a period - it's a retention metric. NRR measures whether existing customers generate more or less revenue over time, accounting for churn and expansion. A business can have 5% monthly churn (strong) but NRR below 95% (weak) if expansion is minimal. Conversely, a business with 8% churn but strong expansion (upsells, add-ons) can have NRR above 100%. NRR is the more important metric for long-term health because it shows whether the business is growing from within.

How do I know if my subscription payback period is acceptable?

Payback period should be 12 months or less. If it's 18+ months, you're funding customer acquisition with future cash flow, which is unsustainable unless you have significant venture capital. Calculate it as: CAC-Sub / (Average Monthly Gross Margin per Subscriber). If CAC-Sub is $100 and gross margin is $10 / month, payback is 10 months - acceptable. If payback is 15 months, either reduce CAC (shift to cheaper channels, reduce discounts) or increase margin (raise prices, reduce product cost).

Should I focus on acquiring new subscribers or retaining existing ones?

Both, but in sequence. First, prove that you can retain a cohort to at least 12 months with churn below 8%. If churn is 15%+, acquisition spend is wasted - you're pouring water into a leaky bucket. Once retention is proven, scale acquisition. The rule: if LTV / CAC is below 3:1, focus on retention and margin first. If it's above 3:1, you can afford to acquire aggressively.

What's a red flag in subscription metrics?

Three red flags: (1) Payback period exceeds 18 months - the unit economics are broken. (2) Churn rate exceeds 12% for two consecutive months - product or pricing is failing. (3) NRR drops below 90% - the business is shrinking from within. If you see any of these, pause growth initiatives and investigate. Growth on a broken foundation is expensive.

Want this on your account?

Thirty minutes. Bring the number that keeps you up.

Book a demo