eCommerce sites running Hyvä on Magento with complex product catalogs — hundreds of filterable attributes, multi-variant products, and rich media galleries — are reporting 3x faster mobile page loads, 40% lower mobile bounce rates, and 31% higher mobile conversion rates compared to identical catalogs on Luma. The data confirms that Hyvä does not just improve simple storefronts — it fundamentally changes what is possible for complex catalog experiences on mobile devices.
The Mobile Performance Crisis for Complex Catalogs
Complex product catalogs create a specific mobile performance problem that simple storefronts never encounter. When your products have 30+ filterable attributes, multiple configurable options (size, color, material, finish), image galleries with 8-12 angles, detailed specification tables, and comparison functionality — the frontend payload explodes.
On Luma, a configurable product page with twelve swatch options, eight gallery images, a specifications table, and cross-sell recommendations generates 3.1 MB of frontend payload on mobile. The page achieves First Contentful Paint at 3.8 seconds and Time to Interactive at 9.2 seconds on a 4G connection. That 9.2-second TTI means the customer cannot reliably interact with size selectors, add-to-cart buttons, or image galleries until nearly ten seconds after tapping the product link.
Google's industry data shows that 53% of mobile visitors abandon a site that takes longer than three seconds to load. For complex catalogs on Luma, the abandonment begins before the page even becomes interactive.
The same product page on Hyvä generates 340 KB of frontend payload. FCP arrives at 0.9 seconds. TTI at 1.6 seconds. The customer is browsing swatches and adding to cart while the Luma version is still loading JavaScript modules.
The Data: Six Months of Production Metrics
Bemeir compiled mobile performance and engagement data from four enterprise Magento implementations that migrated from Luma to Hyvä, each with complex product catalogs exceeding 5,000 SKUs with significant attribute depth. The aggregated results paint a consistent picture.
| Mobile Metric | Luma (6-Month Average) | Hyvä (6-Month Average) | Change |
|---|---|---|---|
| Largest Contentful Paint | 4.6 sec | 1.3 sec | 72% faster |
| Time to Interactive | 8.4 sec | 1.9 sec | 77% faster |
| First Input Delay | 340 ms | 18 ms | 95% reduction |
| Total frontend payload | 2.8 MB | 360 KB | 87% reduction |
| Mobile bounce rate (PDP) | 58% | 35% | 40% reduction |
| Mobile bounce rate (category) | 62% | 38% | 39% reduction |
| Mobile add-to-cart rate | 4.2% | 6.8% | 62% increase |
| Mobile conversion rate | 1.4% | 1.83% | 31% increase |
| Mobile pages per session | 3.8 | 6.2 | 63% increase |
| Mobile session duration | 2:14 | 3:48 | 70% increase |
The most striking data point is mobile pages per session increasing from 3.8 to 6.2. When pages load fast enough for fluid browsing, customers explore more of the catalog. For complex catalogs where cross-selling and discovery drive average order value, this deeper engagement translates directly to revenue.
Filter and Search Performance: Where Complex Catalogs Win or Lose
Layered navigation performance is the make-or-break UX factor for complex catalogs on mobile. When a catalog has 20+ filterable attributes — material, size, color, price range, brand, rating, availability, specifications — the filter interaction pattern determines whether mobile customers can actually find what they need.
On Luma, selecting a filter option triggers a full page reload or an AJAX update that takes 1.5 to 3 seconds on mobile. Each additional filter compounds the wait. Customers selecting three filters experience 4.5 to 9 seconds of cumulative waiting. The data shows that mobile customers on Luma abandon filtering after an average of 1.7 filter selections — they give up before narrowing to relevant results.
Hyvä's Alpine.js-powered filtering completes each filter selection in 150 to 300ms on mobile. The product grid updates while the customer's thumb is still moving to the next filter option. Mobile customers on Hyvä apply an average of 3.4 filters before viewing products — double the Luma average. More filter usage means more relevant product discovery, which drives higher satisfaction and conversion.
One industrial parts retailer with 22 filterable attributes tracked filter-to-conversion funnels before and after Hyvä migration. On Luma, only 12% of mobile visitors who opened the filter panel completed a purchase. On Hyvä, that number jumped to 28% — a 133% improvement in filter-to-purchase conversion. The conclusion was clear: customers were not uninterested in filtering on mobile. They were frustrated by the performance of filtering on mobile.
Image Gallery Performance: Rich Media Without the Penalty
Complex products need rich visual presentation. Industrial components need close-up detail shots. Fashion products need multiple angle views. Configurable products need per-variant imagery. This rich media requirement creates a direct conflict with mobile performance — unless the frontend handles it intelligently.
Luma's product gallery on mobile loads all images eagerly (including thumbnails for the carousel), initializes the Fotorama gallery library (120 KB of JavaScript), and sets up zoom functionality that may never be used. For a product with twelve images, the gallery alone contributes 1.2 to 1.8 MB to the mobile page load.
Hyvä implements a progressive loading pattern. The primary product image loads immediately (priority LCP resource). Subsequent gallery images lazy-load as the customer swipes. Zoom functionality loads on-demand when the customer initiates a pinch gesture. The gallery JavaScript is implemented in Alpine.js — approximately 3 KB versus Fotorama's 120 KB.
The production data confirms the approach. Median product image gallery load time on Hyvä is 0.4 seconds versus 2.1 seconds on Luma. Mobile customers on Hyvä view an average of 4.8 product images per PDP visit versus 2.3 on Luma. The faster gallery interaction encourages the visual exploration that builds purchase confidence — particularly important for products where visual inspection partially substitutes for physical handling.
Configurable Product Performance
Configurable products — products with selectable options like size, color, and material — represent the highest-complexity mobile UX challenge. Each option selection potentially changes the price, availability, images, and description. On Luma, this reactivity runs through KnockoutJS observables that trigger cascading DOM updates.
For a product with three configurable attributes (size with 8 options, color with 12 options, material with 4 options), Luma's configuration requires loading a JSON configuration object of 50-200 KB depending on the number of simple products backing the configurable. Each option selection triggers a recalculation that takes 200-800ms on mobile devices, creating a sluggish interaction pattern that makes customers question whether their selection was registered.
Hyvä handles the same configurable product with Alpine.js reactive data. The configuration JSON loads identically (backend-generated, unavoidable), but the reactive update cycle completes in 15-40ms — fast enough that the interaction feels instantaneous. Swatch images update, price recalculates, and availability refreshes before the customer's finger lifts from the screen.
Bemeir measured the impact on a fashion retailer with extensively configurable products. On Luma, 23% of mobile customers who began configuring a product (selected at least one option) completed the add-to-cart action. On Hyvä, that figure increased to 41% — a 78% improvement in configure-to-cart conversion. The performance difference was the primary variable; the product catalog, pricing, and marketing were identical.
The Cumulative Revenue Impact
The individual metrics each tell part of the story. The cumulative impact tells the business case.
Consider a retailer with $15 million in annual eCommerce revenue, 65% mobile traffic, and a complex catalog averaging $85 average order value. Applying the measured improvements from Hyvä migration against conservative estimates.
Mobile conversion rate improves from 1.4% to 1.83% (31% lift). Mobile revenue increases from $9.75 million to $12.7 million — approximately $2.95 million in additional annual revenue.
Mobile pages per session increases from 3.8 to 6.2. Deeper catalog browsing drives cross-sell revenue, increasing average order value by approximately 8% based on production data — adding another $780,000 in annual revenue.
Combined annual revenue impact: approximately $3.7 million from a Hyvä migration investment of $60,000-$100,000. The ROI materializes within the first four to eight weeks of production operation.
Bemeir presents these projections to enterprise clients during Hyvä evaluation, and the actual results consistently meet or exceed the modeled improvements because the performance differential between Luma and Hyvä is so substantial that even conservative conversion rate lift assumptions produce compelling ROI.
