Storefronts measured on revenue per visitor, not on how the homepage looks in a case study.
Ecommerce is one of the few areas where the effect of engineering is directly measurable. A second of load time, a required account before checkout, a shipping cost revealed on the final step — each has a number attached, and the number is usually larger than people expect.
B2B adds a layer that most storefront themes handle badly: customer-specific pricing, quote requests instead of instant checkout, purchase orders, minimum quantities, and account hierarchies where a buyer orders against a parent company's terms. Forcing that into a consumer theme produces something everyone involved quietly works around.
Shopify when speed to launch and operational simplicity matter. WooCommerce when you need control and already live in WordPress. Headless or custom when the catalogue or pricing logic genuinely exceeds what a platform models.
Customer-group pricing, quote-to-order flows, purchase orders, minimum order quantities and multi-user accounts with approval steps.
Guest checkout, address autocomplete, real shipping costs surfaced early, wallet payments, and recovery for abandoned carts. Most lost revenue is here rather than on the product page.
Image pipelines, lazy loading and script discipline. Third-party tags are the usual culprit, and they are rarely audited once added.
ERP, inventory, accounting and shipping integrations with idempotent webhooks, so a provider's retry does not create a duplicate order.
Moving platforms with URL mapping, 301s, and structured data carried across — the step skipped most often, and the one that costs a year of rankings.
Before new features, look at the funnel you already have. A fix in checkout usually beats a redesign, and costs a fraction.
Variants, bundles, and pricing rules decided up front. Catalogue structure is the hardest thing to change later.
The revenue path gets built and tested before the parts that are more enjoyable to design.
For migrations, every old URL is mapped and tested before the switch, not patched afterwards from a 404 report.
Trade pricing, quote requests, account-based ordering and credit terms.
Custom themes, app integration and storefront work beyond what the theme editor allows.
Moving between WooCommerce, Shopify and custom builds with rankings and order history intact.
A Next.js storefront on a commerce backend, when the standard theme layer is the thing holding performance back.
Shopify if you want to sell rather than run infrastructure, and its model fits your catalogue. WooCommerce if you are already on WordPress and want full control over data and hosting. Custom or headless only when the business logic genuinely exceeds both — unusual pricing, very large catalogues, or a storefront that is really an application. Custom is the most expensive to maintain and worth recommending only when the case is clear.
A good deal, yes — customer-group pricing, quote flows, net terms and account hierarchies are achievable, some natively on Plus and some through apps or custom work. Where it stops being sensible is deeply bespoke contract pricing per customer per product with approval chains, at which point a custom backend is cheaper than fighting the platform. I will tell you which side of that line you are on.
It will if the redirects are an afterthought, and that is the usual cause of a post-migration traffic collapse. Done properly — a complete old-to-new URL map, 301s tested before launch, structured data and metadata carried over, sitemap resubmitted — rankings generally recover within weeks. The risk is real but it is a process problem, not an unavoidable cost.
Yes, on a retainer: platform and dependency updates, security patches, performance monitoring and small changes. Ecommerce sites decay faster than most because they accumulate apps and tracking scripts, and nobody owns removing them.
That is often better value than a rebuild. It starts with an audit of the funnel, page speed and checkout, producing a ranked list of fixes with the expected impact of each. You can then have me implement them or hand the list to your own team.
A short call costs nothing, and if your project is a better fit elsewhere I will say so. Replies usually within 24 hours.
Get in touch