Zubr, Wholesale Meat Products Marketplace

Client Journey

Wholesale Buyer

1. Arrives with a specific need. A HoReCa manager or retail chain buyer lands on the platform looking for a particular meat category, chicken, beef, or duck. The homepage immediately surfaces the right catalog section with no detours.
2. Browses a structured catalog. Products are organized by type and origin (Belarusian producer, local distributors). Each card shows cut type, packaging format, and temperature regime, refrigerated or frozen. No ambiguity, no need to call for specs.
3. Reads about quality guarantees. A dedicated section explains the supply chain: animals raised without antibiotics under nutritionist supervision, all documentation provided by the supplier. The buyer builds trust before any contact.
4. Submits a bulk inquiry in 90 seconds. The inquiry form captures name, phone, and product interest. The manager receives a WhatsApp ping and responds within 5 minutes during business hours, a guarantee published on the site.
5. Gets a tailored commercial offer. The team discusses volume, assortment, and delivery schedule. A contract and prepayment follow, then production, quality control, and delivery to the buyer's warehouse or point of sale.

Content Manager (Internal)

1. Logs into Umbraco CMS. The back office is role-separated, catalog editors, page editors, and admins operate in their own sections with no risk of cross-contamination.
2. Adds or updates a product. A structured form enforces consistent data, category, cut type, origin, certificates. No free-form HTML; every field is typed.
3. Publishes instantly. The change propagates via GraphQL to the Next.js frontend. The index engine picks up the update and re-warms the cache; the page reflects the new data in seconds without a full rebuild.

Before & After

Metrics

Architecture Highlights

• Headless CMS (Umbraco + .NET Core). Content is modelled as typed documents, with no free-form HTML in the back office. Editors work in a structured form; developers consume a clean API.
• GraphQL data layer. The frontend queries only the fields it needs per page. A single unified schema covers products, categories, pages, and settings, with no REST endpoint sprawl.
• Index engine + request caching. All GraphQL queries on the frontend pass through an index service that caches responses in memory. Cache invalidation fires automatically on CMS publish events, stale content is never served.
• React SSR via Next.js. Pages are server-rendered on first load for instant paint and correct SEO indexing. Client-side navigation is handled by React's virtual DOM, with no full-page reloads, zero visible latency between routes.