Knowledge Base
Knowledge Base » Capabilities » E-Commerce Operations – Running Live Stores, Not Just Building Them
📝 Context Summary
E-Commerce Operations
Building a store and operating one are different jobs, and this is the second one. Product-building is the build – the catalog, the buyer tools, the funnel. E-commerce operations is what happens every day after launch, where the enemy isn’t an ugly page but an inaccurate one: a price that’s stale, an item that sold last week still sitting there for sale. I operate several live WooCommerce stores, and the skill is keeping them correct while they’re genuinely different kinds of store.
The first rule is that stock and price are never allowed to lie. In operations the failure mode is quiet. A store that shows something in stock when it’s gone, or yesterday’s price today, doesn’t crash – it just burns trust one order at a time and generates refunds and apologies. So the operating discipline is non-negotiable: WooCommerce is the single source of truth, product data is kept current, and anything sold is pulled before it can be offered again. It’s the same rule the Bernard Hats Shop Assistant enforces for the AI layer, but it starts with the humans and the catalog long before any bot is involved.
Unit-of-one inventory changes how you run the whole store. Bernard Hats isn’t normal retail – every piece is a quantity-of-one vintage item, so “sold” means gone, not “reorder.” That single fact drives the operating model: products are pushed once and then the live store is authoritative, sold pieces follow a delete/retire convention instead of lingering at zero stock, and every attribute – brand, color, size, material – is a canonical term from a controlled vocabulary, not free text a listing invents. Let attribute terms sprout freely and the catalog’s filters rot within a month; disciplined taxonomy is what keeps a one-of-a-kind store navigable.
Different stores are different commerce models, and I run several at once. The operating skill is knowing which model a store is and not forcing one pattern onto another. Bernard Hats is unit-of-one vintage. top-pizza.com is a configurable menu, where sizes and toppings run through TM EPO Global Forms deliberately instead of WooCommerce variations – a completely different product mechanic – and I operate the commerce and product layer (sync, catalog, recommendations) while the client runs the restaurant itself. PIVIT and Stubbe’s are B2B spec catalogs with no impulse cart at all – the “purchase” is spec sheets, a volume calculator, and routing to the nearest dealer. Same platform, three incompatible definitions of “buy,” operated correctly side by side.
The storefront isn’t the only shelf. Channel choice is an operations decision, not a marketing afterthought. Bernard Hats’ clearance inventory moves through eBay and Facebook Marketplace alongside the Woo store, and each channel has its own rules – velocity over margin when the goal is to move stock, listing conventions that differ per platform (Marketplace listings drop the word “Vintage,” for instance), and a presentation standard that doesn’t get cut just because it’s a secondary channel. And marketplace operations go deeper than clearance: I’ve run real Amazon selling – for the Rejuvenate skin-care brand and during my time at Sparta – as a full discipline, not a listing dump. That’s meant extensive listing work, Amazon advertising campaigns, and Brand Registry on both, plus FBA (Fulfillment by Amazon) logistics at Sparta. Amazon is the hardest version of the truth problem, because you’re keeping catalog data accurate and compliant on a platform whose rules you don’t control – and its own advertising and fulfillment machinery to operate on top of that. Running multi-channel means the same product is accurate everywhere it’s listed and disappears everywhere the moment it sells.
What I’ll own honestly. This is small-catalog, hands-on operation on WordPress and WooCommerce plus the strongest tools in that ecosystem – not enterprise commerce. I don’t run Shopify Plus at scale, an ERP-backed ten-thousand-SKU operation, or a warehouse management system, and I won’t pretend to. My edge is narrower and real: keeping several small, live, genuinely different stores accurate and moving at once, myself – on the same fleet I operate everything else on, with the search discipline that makes the products findable. Operating a store well isn’t glamorous. It’s just never letting it lie.