E-commerce built so the catalogue ranks and the checkout converts.

By your economics, not my preference. Catalogue size decides how much of the store can be generated at build time and how much has to stay dynamic. Traffic shape decides where caching sits. Subscriptions, complex pricing, B2B rules and unusual fulfilment decide where the service boundaries go. I write that comparison down, including costs of change, ownership and integration risk, before anything is designed, so you can disagree with it.

You do, in full, from the first commit. Code lives in your repository, deploys from your hosting account, and connects to gateway and analytics accounts in your name. There is no licence, no hosting lock-in and no agency-owned wrapper. If you decide to move to another developer, you hand over access and everything keeps working. That is the test I build against.

The four I have integrated in production are N-Genius, Adyen, Bank of Muscat and PointsPay, all on the same booking flow. For a UAE store N-Genius and Adyen are the usual pair. Elsewhere the provider changes but the integration work does not. The integration matters more than the logo: webhooks, 3-D Secure, refunds, idempotency and reconciliation. I also flag onboarding times early, because merchant approval takes days with some providers and weeks with others, and that timing shapes your launch date.

That is the default requirement, not an extra. I crawl every existing URL (products, categories, filtered pages, discontinued lines) and map each one to its destination before rebuilding starts. Metadata, canonicals and Product schema carry over or improve deliberately. I did exactly this on Thrifty's PHP Laravel to Next.js migration in the UAE with no ranking drop, and the same playbook applies to a catalogue.

Six to twelve weeks from signed scope to live checkout, depending on catalogue size, number of integrations and how ready your product data is. Discovery and the platform decision take the first week. You see a staging URL every week after that, so progress is visible rather than reported. Gateway merchant approval runs in parallel, because it is usually the longest external dependency.

Thirty days of monitoring are included: indexation, Core Web Vitals, gateway errors and purchase-event accuracy, reported weekly in writing. Stores break in specific ways — a failed webhook, a schema field that stops validating, a filter page that starts getting indexed — so the watch window is aimed at those. Beyond thirty days, optional ongoing support keeps the same checks running.

Often. I am one developer, not an agency, so I slot into your workflow rather than replacing it: your repository, your branch strategy, your ticket board, your review process. A common split is that I own storefront, checkout, schema and tracking while your team keeps the backend or ERP integration.