Growing a subscription business: channels, sequencing, and unit economics

Executive overview

Most subscription startups over-rely on paid acquisition or chase diversification too early. The right growth channel depends on your LTV, product structure, and stage — not on what's trendy.

Pick the channel your product is built for, prove it works, then diversify — not the other way around.

The three growth engines and when to use each

• Virality and referrals work when the product has inherent network effects — utility rises with more users (collaboration tools, marketplaces). Hard to manufacture if absent from day one.
• Beloved brands can run incentivised referral programs even without network effects.
• SEO requires at least two of: a unique editorial angle, a programmatic data angle, or a unique data asset. Payoff horizon is 6–12 months minimum.
• Paid acquisition requires strong LTV (hundreds of dollars, not $50–60), free-to-paid conversion above 5%, and healthy funnel metrics before scaling.
• Grammarly leaned on performance marketing; Canva built on long-tail programmatic SEO — neither diversified until one engine was fully built out.

Sequencing: when to focus vs. diversify

• The 80/20 rule applies even at late-stage companies — one channel usually drives the majority of growth.
• Early-stage mistake: worrying about over-reliance before the primary engine is at scale. Fix: keep pouring fuel on what's working.
• Late-stage mistake: 90%+ reliance on a single channel at $50M+ ARR. Fix: carve out bandwidth now to explore alternatives.
• Focus paired with rapid iteration beats both extremes — pure focus risks wasting runway on the wrong bet; spreading thin means nothing gets a real shot.
• Abandoning a channel doesn't mean never revisiting — re-evaluate with fresh eyes as the company's position changes.

Paid acquisition: what's changed

• Short-term contraction in paid budgets is real; the payback window expectation has tightened from 12 months to 6 months or less.
• iOS 14 and cookie deprecation eroded click-based attribution — it was always correlative, never causal.
• Media mix modeling (Recast) is making a comeback for companies spending $100K+/month across 3+ channels.
• Incrementality testing (Measured, Incremental) is the only way to establish causal attribution — controlled experiments, not last-click models.
• Companies with strong unit economics (Grammarly, Canva) benefit from contraction: competitors pull back, CPMs fall, less auction competition.
• Two valid reasons to run paid: (1) efficient acquisition at scale, (2) fast learning on messaging and positioning. Know which camp you're in.

SEO: when and how to invest

• Series A companies are now exploring SEO earlier because paid is less attractive and runways are longer.
• Three-month time-box is a useful rule of thumb to decide if SEO is likely to work before committing further.
• An external audit (boutique agency or solo consultant) costing $5–10K is faster and cheaper than building internally from scratch.
• Ethan Smith at Graphite is a recommended resource for first-principles SEO thinking.

Conversion and onboarding benchmarks

• Healthy website-to-free-account conversion for prosumer freemium: 20–35% at scale.
• Free-to-paid conversion below 5% is not viable long-term regardless of top-of-funnel volume; north of 7% is solid.
• Onboarding is almost always a high-ROI investment — early-stage companies can 2–4x activation rates; Series B+ companies can still see 20–30% lift.
• The central onboarding challenge for complex products: democratise access without overwhelming with the full feature set at once.

Offline and emerging channels

• TikTok reaches broader demographics than founders assume — including 40+ high-income audiences — and remains less saturated with advertisers than most digital channels.
• Out-of-home, podcasts, and direct mail are getting a second look as online attribution has deteriorated and offline attribution has improved.
• The attribution gap between online and offline has narrowed significantly, making offline ROI more measurable than it was five years ago.

What always pays off vs. what bites back

• Customer research — interviews, surveys, panels — consistently unlocks clarity and team alignment, especially at Seed to Series B.
• Over-reliance on paid acquisition creates structural risk; companies that don't diversify while at scale are exposed to algorithm and policy changes.
• Attribution investment has a right level for each stage — under-investing leads to spending inertia; over-investing chases precision that doesn't pay off at small budgets.