Speed is revenue in retail, and that is not a slogan — it is the one sector where the engineering decision and the marketing number are the same decision. A slow product page loses the sale the ad paid for.
Where the money leaks
Four places, in the order we usually find them: page weight on category and product templates, a checkout with a step nobody needs, stock that says “available” when it is not, and a tracking setup that quietly lost half its events to a browser update.
Fixing the fourth one first is tempting because it is cheap. Fixing it last is correct, because the other three change the numbers it reports.
Catalogue, feeds and the boring layer
Product data is the platform. Variants, pricing rules, availability by warehouse, and a feed that a marketplace or an ad network will accept without manual cleanup every week. Most retail projects that feel slow are not slow at the framework level; they are slow because the catalogue query was never designed.
Integrations follow the same rule as the rest of our integration and API work: one place owns the truth, everything else subscribes.
Tracking that survives the browser
Server-side events, consent handled properly, order values reconciled against the actual ledger and not against a pixel’s optimism. When the finance report and the ad platform disagree, we want to know which one is wrong before the budget meeting.
This is the same discipline the analytics and CRO track runs everywhere; retail just punishes getting it wrong faster.
Demand, channel by channel
Feeds and paid social do the volume. Search does the margin — category and buying-guide pages compound while campaigns stop the day the card is declined. Email and lifecycle work is the cheapest revenue in the stack and usually the most neglected; see email and CRM.
Seasonality is planned around, not reacted to. If the year is decided in six weeks, the platform is load-tested in month eight and the creative is finished in month nine.
Where models earn their place
Search relevance, merchandising and support deflection. AI in e-commerce and retail sets out what we automate and what stays a human decision.
Other sectors, and what changes between them, are on the industries page.