Web · Improvements · E-commerce · UX · Performance
Almamed online store — ongoing improvements
Since 2021 we run the Webasyst store: catalog UX and path to order, homepage upgrades (banner, trust blocks), then a large 2026 speed pool — search, N+1, cache, and Core Web Vitals.
About the client
Almamed supplies medical equipment nationwide: direct contracts with manufacturers and a wide range. Site on Webasyst: first we rebuilt template UX and the main storefront, then — as the catalog grew — we fixed search lag, listing weight, and Google metrics.
The pain they came with
Two stages. First the client came with “the store exists, but buyers get lost”: silent product cards, weak search, important buttons buried in the template, homepage not selling trust. Later, as assortment and traffic grew, a second pain appeared — heavy pages: catalog and search took seconds, brands loaded half a megabyte of HTML, PageSpeed went red on LCP/CLS/TTFB.
-
Card and catalog don’t sell
No SKU, weak specs, awkward filters — hard to compare medical-equipment models.
-
Homepage doesn’t hold attention
The banner “jumped” on load; trust and assortment blocks didn’t cover the B2B “understand first, then buy” path.
-
Search and metrics on fire
SearchPro slowed the storefront, brands returned hundreds of KB of HTML, LCP/CLS/TTFB hit SEO and ads.
How we guided them
First we fixed the path to order and the homepage storefront. When the store “learned to sell”, we took on speed: a 30+ task checklist, SearchPro rewrite, and live PageSpeed measurements — without stopping the storefront.
- 01
Stage 1 · UX: path search → card → cart → delivery
- 02
Rebuilt header, cards, filters, delivery, and catalog backend
- 03
Upgraded the homepage: banner slider, trust blocks, popular products
- 04
Stage 2 · speed (2026): N+1, SearchPro, brands, sitemap, fonts, CLS/LCP
- 05
Removed extra SQL in product lists, SEO names, and subcategories
- 06
Rewrote SearchPro from inside: zero-cost field, suggest/page cache
- 07
Pagination on /brands/ (was ~547 KB HTML on one screen)
- 08
Slider FOUC, retina banners, deferred chat widgets — still ongoing
In plain speak
Brief jargon — without the fog. How we explained each step to the client.
Path-to-order UX
Not “pretty buttons” — how many steps to “buy” without hunting for a phone in the template.
Banner FOUC / retina srcset
The homepage slider used to flash and load everywhere. Now — homepage only, first frame immediately, Retina images without blur.
N+1 in the catalog
Before: a separate DB query per product. After: one batched query. Lists open noticeably faster.
SearchPro v2 / zero-cost field()
The search field no longer slows every page. The shell is light; smart work runs only when someone starts typing.
LCP / CLS / TTFB
Three Google numbers: how fast main content shows, whether layout jumps, how fast the server answers. On the homepage LCP and CLS are in the green.
What we did, step by step
Each step is a concrete change. Meaning on the left, screen on the right.
Homepage that sells the storefront
Live homepage: header with search and cart, banner slider without a “stack” on load, trust block (since 2020, 300+ clinics, 18,000+ SKUs) and popular products. Retina banners and height reserve — so it doesn’t jump on mobile.
Header and menu — everything at hand
Phone, cabinet, cart, search, and delivery always available. Buyers don’t hunt for “where to click”.
Catalog and filters
Subcategories, quick brand filters, sort and view modes. Lists sped up: removed N+1 and added lazy previews.
Product card that answers questions
SKU, specs, price, and quick actions. Fewer “what model is this?” calls — more ready orders.
Search that doesn’t slow the storefront
SearchPro rewritten from inside. Category field is almost free for the server. Suggest and search page — cached. On mobile — one field instance.
Brands and speed for Google
Manufacturer catalog with pagination instead of a giant page. Plus guest HTML cache, fonts, CLS/LCP. Field mobile: homepage LCP 2.1s, CLS 0.02.
More frames from the project
Not a gallery for its own sake — a few screens that complete the picture.
Where we landed
First the store became usable: path to order, clear homepage, cards and delivery. Then — fast: search, catalog, and Google metrics stopped being the bottleneck. Fresh screenshots — from live almamed.su after the 2026 pool.
What they got
- UX since 2021: header, cards, filters, delivery, backend
- Homepage: slider without FOUC, retina banners, trust blocks
- Removed N+1 in product lists, SEO names, and subcategories
- SearchPro v2: light field, suggest and search-page cache
- /brands/: ~547 KB → ~169 KB via pagination
- PageSpeed field: homepage LCP 2.1s · CLS 0.02 (mobile)
- Lab desktop CLS: 0.6 → 0
- Store on support — improvements continue
In numbers
Need an online store or improvements?
We’ll rebuild the catalog, homepage, and path to order — and if the store is already heavy, we’ll tackle speed as systematically as we did for Almamed.
Other projects
Similar cases and neighbouring directions
15,000+ products in catalog
VILMED
VILMED storefront improvements
We run a large medical-equipment store on Bitrix: sticky header and smart search, catalog and product-card UX, SEO filters, ERP sync, composite/Brotli, and shell cleanup. Live — vilmed.ru.
View case →
55,000+ SKUs in the catalog
KosmaMed (formerly MedMarket)
KosmaMed online store
We run a large medical-equipment store on Bitrix: MedMarket→KosmaMed rebrand, domain migration with 301s, speed (no RAND, composite, WebP), MoySklad/ERP/carriers. Live — kosmamed.ru.
View case →
~1,900 SKUs in the catalog
Oftalmag
Oftalmag online store
We run an ophthalmology-equipment store on Bitrix (enext): Heine/KaWe/Riester catalog, “request a price”, city picker, personal-cabinet SEF. Live — oftalmag.ru.
View case →