Architecture MOC

Date: 2026-08-16

How the parts are arranged, and how a change gets from a laptop to a customer without anyone holding their breath.

Rendering and delivery shape

The decision that constrains most others.

Rendering Strategies — CSR, SSR, SSG, ISR, streaming, islands: what each trades
Hydration — making server-rendered markup interactive, and why it’s the expensive part
Islands Architecture — hydrating only the interactive bits
Edge Computing — running logic at the CDN, and the constraints that come with it
Static Generation and Revalidation — build-time rendering, and keeping it fresh without rebuilding everything
Headless Architecture — decoupling the front end from the commerce or content backend
Composable Commerce — assembling a stack from specialist services, and the integration cost that arrives with it

Monolith vs Services — the actual tradeoff, which is organisational before it’s technical
Backend for Frontend — a BFF: one API shaped for one client, rather than one API for everyone
API Design — resources, naming, versioning, errors, pagination
REST GraphQL and RPC — three models, and the client-server relationship each assumes
Webhooks — inbound events from someone else’s system, and the reliability problems they hand you
Message Queues — decoupling producers from consumers, and delivery guarantees
Event-Driven Architecture — systems that react to facts rather than call each other
Multi-Tenancy — one system serving many customers, and the isolation decisions
Integration Patterns — point-to-point, hub, iPaaS, and how the mess accumulates
Building with Language Models — a non-deterministic, high-latency, metered dependency; retrieval, grounding, and the failure modes that are new

Feature Flags — decoupling deploy from release; the single highest-leverage practice here
Progressive Delivery — canary, percentage rollout, ring deployment
Blue-Green and Rolling Deployments — two ways to swap versions without downtime
Environments — local, preview, staging, production, and the drift between them
Database Migrations — changing schema without downtime, and the expand-contract pattern
Rollback and Forward Fix — deciding which, before you need to

Technical SEO is architecture, not marketing. The keyword and content half lives in Acquisition.

Technical SEO — the durable mechanics, separated from the parts that move with each algorithm update
Crawling and Indexing — how a crawler discovers, fetches and stores pages, and the difference between the two verbs
Rendering and SEO — what crawlers execute, and the cost of client-side rendering
URL Structure — designing it once, because changing it is one of the most expensive things a site can do
Canonicalisation — duplicate URLs and telling a crawler which one counts
Redirects and Link Equity — 301 versus 302, chains, and what a migration loses
Faceted Navigation and Crawl Budget — filter combinations generating infinite crawlable URLs
Structured Data — product, review and breadcrumb markup, and what it’s actually eligible for
Site Migrations and SEO — the highest-risk routine event in ecommerce

The Strangler Pattern — replacing a system incrementally by routing traffic away from it piece by piece
Technical Debt — the useful definition, and distinguishing it from code you dislike
Refactoring Strategy — changing structure without changing behaviour, in steps that stay shippable
Deprecation — removing something people depend on, with notice and a path
Replatforming — the ecommerce-specific version, and the measurement continuity problem it creates

Observability — logs, metrics and traces as three distinct questions
Error Tracking — capturing, grouping and prioritising what breaks
Alerting — alerting on symptoms rather than causes, and alert fatigue
Service Level Objectives — SLIs, SLOs, error budgets, and what they’re for politically
Incident Response — the sequence, and blameless postmortems as the actual deliverable
Rate Limiting — protecting a system from its consumers
Graceful Degradation — deciding in advance what stops working first

Filed in other domains, needed constantly here.

Foundations — Idempotency · Backwards Compatibility · Secrets Management · Coupling and Cohesion · Eventual Consistency · CAP Theorem
Tooling — Branching Strategies · Supply Chain Risk · Environment Configuration · Preview Environments · Continuous Deployment
Experimentation — Client-Side vs Server-Side Testing · Rollouts as Experiments
Analytics — The Data Layer · Event Taxonomy Design · Tag Managers
Commerce & Growth — Replatforming
Platforms — Shopify · Vercel · Cloudflare