Substack reshaped paid newsletter economics from 2017 onward by giving writers and publishers a zero-upfront-cost path to launch paid subscriptions, in exchange for a 10% share of all subscription revenue. The deal is genuinely good for publications in the early bootstrap phase where the operational lift is the bottleneck and revenue is small. As publications grow the maths shifts — at £80,000+ annual subscription revenue, the 10% share becomes £8,000+ of annual cost, and the question becomes whether the platform features still justify the share.
What Substack is good at
Three things Substack does materially better than the alternatives at small scale. The bootstrap path is genuinely zero-cost — no upfront subscription fee, no infrastructure cost, no integration cost; the 10% revenue share is taken only on actual paid subscriptions. The audience-discovery features (Substack Network, Notes social feed, recommendations from other Substacks) drive meaningful new-subscriber growth that flat-fee platforms (Ghost, Beehiiv) cannot replicate at the same level. The editorial pipeline is clean and opinionated — minimal configuration overhead, fast publishing, members-and-payments built-in.
What Substack is not good at
Three things Substack does worse than focused alternatives. Cost at scale — the 10% revenue share scales linearly with revenue, so a publication growing to £150,000 annual revenue pays £15,000/year to Substack alone, which buys substantial alternative infrastructure many times over. Customisation flexibility — Substack’s minimal customisation surface suits early-stage publications but constrains established publications wanting brand-distinct presentation, custom commercial functionality, or service-and-product pages alongside the newsletter. Schema depth — Substack emits decent Article schema by default but does not expose the deeper editorial schema (citation arrays, mentions, audience entities, AI-crawler fields) that lifts AI-search citation in 2026.
When Substack is the right answer
Two scenarios where staying on Substack is the better economic and operational call. First: bootstrap phase — you have not yet validated paid-subscriber market fit and the 10% share is the cost of avoiding infrastructure investment until you do. Pay Substack while you find product-market fit. Second: discovery-dependent publications where Substack Network, Notes and recommendations drive a meaningful percentage of new subscribers, and that audience-acquisition channel is hard to replace.
When the migration is overdue
Three patterns that point to migration. The publication has crossed £50,000+ annual subscription revenue and the 10% share is now meaningful in absolute terms (£5,000+/year). The publication has grown into needing commercial functionality Substack is not built for — service pages, e-commerce, location pages, custom integrations. The customisation constraints are limiting brand presentation in ways that affect conversion or retention.
The cost comparison at common revenue tiers
£30,000 annual subscription revenue: Substack 10% share = £3,000/year. Ghost(Pro) Creator at 1k members: £300/year. Custom Stripe-direct build + Ghost-or-equivalent: ~£300-£500/year ongoing. £80,000 annual subscription revenue: Substack share = £8,000/year. Ghost(Pro) Creator: £300/year. Custom build: ~£500-£800/year ongoing. £150,000 annual subscription revenue: Substack share = £15,000/year. Ghost(Pro) Team: £720/year. Custom build: ~£800-£1,500/year ongoing. The maths gets dramatically harder for Substack as revenue grows.
The migration destinations in practice
Three destinations cover most Substack migrations. To Ghost(Pro) — for publications wanting Substack-equivalent editorial pipeline at flat-fee pricing rather than revenue share. The trade is editorial-experience continuity for the 10% recovered margin. To Beehiiv — for growth-focused publications wanting stronger analytics, ad-network monetisation, and active growth tooling. The trade is similar to Ghost on cost structure but with different feature emphasis. To a custom Stripe-direct build — for established publications wanting full ownership, lowest sustainable cost, and the flexibility to build commercial functionality alongside the publication.
The migration sequence
Content export via Substack’s built-in export tool — Markdown for all published posts, CSV for the subscriber list including subscription status and Stripe customer IDs. Destination platform setup (Ghost, Beehiiv, custom build) with custom domain configured. Subscriber import with Stripe customer transfers for paid subscribers (no re-billing). 301 redirects from Substack URLs to new URLs. First few editorial posts cross-published on Substack for one or two cycles before fully moving. Substack subscription cancellation after the migration is verified. Most under-10,000-subscriber publications complete the migration in 2-4 days of structured work.