Stripe is live. The hard part starts now: metering, proration, dunning, tax, reconciliation.
Billing engineering, not "we'll install Stripe." Webhook reliability, metering, proration, dunning, tax, multi-currency, reconciliation. For teams already live on Stripe who need it right.
Engagement scope · Draft
#BILL-2026-001- Senior Billing Engineering
- Webhook reliability harness
- Usage-based cost model
- Migration playbook (WCS to Stripe to Zoho)
Numbers from the work
Subscriptions generated on PlusHeat (Zoho)
Handled across billing implementations
Per-user/month modelled cost (Mux baseline)
Off-for-3-months promo running cleanly
What you get
Six engagements. Each one a separate billing discipline.
Stripe Billing builds
Checkout, Customer Portal, Tax, Connect for marketplaces. Real plan/coupon/promo logic, not a Stripe Checkout drop-in.
Usage-based & metered billing
Including modelling the unit economics before the billing schema is locked. Per-user, per-event, per-minute, per-MB, whichever the product actually charges on.
Subscription migrations
Stripe, Zoho, WCS, Chargebee, Recurly. Renewal cadence preserved, payment methods migrated where the gateway permits, proration state honoured.
Webhook reliability
Idempotency keys, replay handling, retry, dead-letter queues, audit logs. The boring part that costs the most money when it breaks.
Dunning engineering
Smart retry windows, customer-friendly recovery emails, store-credit application, fallback charge methods. Recovers the renewals others write off.
Reconciliation
Stripe events to application database to accounting (Xero / QuickBooks / Zoho Books). End-of-month closes in hours, not days.
Per-user cost curve
Three pricing models. One curve tells you which one you actually have.
Modelled per-user cost across 0 to 10,000 users for flat-fee, tiered, and usage-based structures. The Mux-derived usage curve plateaus at $4.42/user/month at scale, unrecoverable if your pricing was set assuming flat.
Flat fee
$7/user/month baseline, declining slightly with scale.
Under-charges past 2,000 users.
Tiered
Step function: $8 under 1K, $5.50 to 5K, $4.20 beyond.
Over-charges between 500 and 1,000.
Usage-based (Mux)
Grows then flattens after the free tier exhausts.
Stabilises only past 5,000 users, at $4.42/user/month.
Inflection: flat-fee under-charges past 2,000 users; tiered over-charges between 500 and 1,000; usage-based stabilises only past 5,000.
Evidence
Three engagements. Three different billing disciplines.
Combinatorial matrix · 40K+ subs
PlusHeat: Zoho orchestration
Plan x homeowner/landlord x call-out excess x monthly/annual x add-ons x promo x DD/card. Orchestration layer between marketing site, customer, and Zoho: source-of-truth discipline, idempotent webhook handling, promo stacking that doesn't leak revenue.
Store credit · SQL gotchas
INSPIRELLE: store credit recovery
Mapping coupon amounts back to renewal totals via raw SQL: wp_postmeta where post_title = coupon_code joined to meta_key = 'coupon_amount'. Resolved MySQL collation errors (#1267, illegal mix of collations) on the join. Real engineering, not a plugin install.
Cost modelling first
Player Profile Hub: Mux economics
Modelled $0.003/min storage, $0.0008/min delivery blended, 100K free minutes/month, 75% watch-rate cap, 50/50 720p/1080p blend (0.9 factor). Baseline: $4.42/user/month at 10 views/min. Delivered as Word report plus interactive React calculator on Vite + React 18 + Tailwind.
Pair this with WooCommerce Subscriptions (the recovery layer underneath), SaaS web app development, and API and integration work for the webhook and reconciliation infrastructure.
Good fit if
When billing is real engineering
- Teams with Stripe live who need the next layer: metering, proration, multi-currency, reconciliation
- Products with usage-based pricing where the unit economics need modelling before billing ships
- Operators migrating between billing platforms without losing renewal cadence
- Engineering teams who want webhook reliability done by someone who has shipped it three times
Probably not a fit
When the Stripe docs are enough
- First-time Stripe Checkout install with no subscription, no metering, no tax complexity. Read the Stripe docs
- Stores under $100K ARR where the engineering cost outpaces the recovered revenue
- Teams looking for someone to copy-paste the Stripe quickstart
Tools we use
Stripe by default. The others when the buyer made that decision already.
- Billing
- Stripe Billing · Customer Portal · Tax · Connect · Issuing where relevant
- Alt billing
- Zoho Subscriptions · Chargebee · Recurly · WooCommerce Subscriptions
- Reliability
- Idempotency keys · replay queue · DLQ · structured audit log
- Accounting
- Xero · QuickBooks · Zoho Books reconciliation
If Stripe is live and something's off
Bring the failure mode. We'll diagnose inside a call.
More in E-commerce & Payments
Web & SaaS
Mobile Apps
E-commerce & Payments
WordPress
CMS & Platforms
37 services across 7 practice areas.