Custom subscription infrastructure for Shopify.
For when Seal, Recharge, and Appstle aren't enough — and full-custom is too much. I build on the subscription APIs directly: custom portals, custom rules, custom backends, for brands whose program has outgrown an app's settings page.
Seven years running The Cumin Club's custom subscription platform · in production since 2019
Subscription apps are excellent — until your program stops being standard.
Build-a-box. Prepaid gifting. Custom cadences and swap rules. A migration that has to preserve your logic. Dunning tuned to your margins. At that point support tells you "you'd need a developer for that."
I'm that developer.
The full customer journey, owned end-to-end
Branded customer portal, custom cancellation flows with save offers, pause / skip / swap logic, account management, win-back — built on Seal's APIs. Seal's native portal gives you theming, toggles, and a built-in cancellation flow, and it documents a JavaScript API for overriding that flow entirely. My work is the distance between configuring a portal and owning the journey.
Cadence logic, swap rules, prepaid and gifting flows — built on Seal's API.
Move platforms with custom logic, discounts, and subscriber history intact.
Retry logic and recovery flows tuned to your margins.
GA and Shopify Analytics instrumentation, funnels built, and the numbers your app's dashboard won't surface.
Rapid, AI-assisted theme development shipped through Shopify rollouts.
Seven years operating one brand's subscriptions — on both sides of build-vs-buy.
Started fully custom
The Cumin Club's ordering platform began as a ground-up build — Next.js and Stripe, written by me. Every rule the program needed, exactly as specified. And every webhook, billing edge case, and retry path mine to keep running.
Why we migrated — deliberately
Over the years we moved the program onto Shopify and Seal Subscriptions — for the ecosystem, not the settings pages. The tools the brand wanted, retention to reviews, integrate first-party with Shopify, and the founder trusted an ecommerce-tested checkout to convert better than a Stripe build. The migration went in with the program's logic intact. I've moved a brand off custom, not just onto it.
Custom on Seal from day one
Seal runs the billing. The parts subscribers and the team touch stayed custom, built on the APIs: account management, admin infrastructure the team runs the program from, retention logic Seal doesn't ship. The standard portal was never part of the build.
Operating it
Ongoing conversion and churn experimentation. Analytics instrumentation across GA and Shopify Analytics, down to the funnels. And an AI-assisted theme-iteration pipeline — variants drafted with Claude, shipped and measured through Shopify's native rollouts.
What seven years teaches
Judgment about what deserves to be custom. Most of what looks like a build problem is a boundary problem: knowing what belongs in the app, and what belongs in code you own.
[TK: TCC testimonial — words only, no numbers]
I've built fully custom billing. Most brands shouldn't.
The best stack is the least custom one that fits.
If your portal looks like everyone else's, so does your retention.
Custom infrastructure needs an owner — the care retainer is the default follow-on. A growth & iteration tier (testing and CRO on the custom stack) is available · case study [TK: coming]
Solo, by design. Seven years of Shopify subscriptions in production. I do the work personally — direct access, no account managers, no handoffs. And I'll sometimes advise less custom than you came in wanting; an agency is structurally incentivized never to.
Send me your store. I'll find 2–3 revenue leaks.
A short recorded review of your live subscription flow — what's leaking, and what it would take to fix. No pitch attached.
Got it — the teardown is in the queue.
I reply personally.