Performance — does Hyvä really matter for catalog-heavy sporting goods?

For sporting goods, more than most verticals. Three reasons:

• PDPs are image-heavy and variant-heavy — 6–12 product photos per PDP plus 80–200 variant SKUs (footwear) or 60–120 (apparel) is normal. Luma + jQuery + RequireJS load all variant data eagerly, killing LCP. Hyvä uses native lazy-load + Tailwind aspect-ratio classes + responsive image srcset + Alpine reactive state for variant pickers, holding LCP under 1.5s on mobile.
• Fitment quizzes need fast interactivity — 8–14 short questions, instant transitions between steps. Luma re-runs the full PDP JS on each step (200–400ms INP). Hyvä uses Alpine.js (under 80ms INP). Quiz completion rate is 22–35% higher on Hyvä in my data.
• Seasonal traffic spikes — 4–12x baseline during first-snow weekend / Black Friday / back-to-school. Hyvä holds; Luma queues. The TTFB difference is what keeps the site up vs crashing in the first 30 minutes of a sale.

Numbers from a recent Luma → Hyvä migration on a 4,500-SKU cycling + outdoor store: Lighthouse mobile 38 → 95, LCP 4.6s → 1.2s, INP 410ms → 68ms. Measured conversion lift: +22% on mobile, +13% on desktop. Migration timeline: 6 weeks. ROI inside 90 days. Bonus: Black Friday survived 8x baseline traffic with origin CPU under 60%.

If you’re on Luma + 3 personalization apps and serious about ranking on category head terms (e.g. “trail running shoes”, “all-mountain skis”, “road bike size guide”), Hyvä migration is the highest-leverage spend you can make.