Activewear 3-Piece Try-On: Bundle Sports Tops, Leggings & Zip-Through Jackets Privately

Architect the 3-Piece Active Set with custom.pair_with Metafields

Map every sports top to its matching leggings and zip-through jacket with Shopify metafields so Style Swap AI layers a complete studio set in a single session. Pair at the variant level—colorway and size—so a shopper who selects a Sage sports bra in S sees the Sage high-waist legging and jacket rendered to match. When pair_with is empty, fall back to the Shopify Product Recommendations API or DOM cross-sell detection.

• Define custom.pair_with as a JSON array of up to 3 ProductVariant GIDs per product (top → bottom → outerwear); variant-level pairs keep colorway and size accurate across the rendered set.
• Bulk-sync metafields via the Shopify Admin bulk editor or the GraphQL productUpdate mutation; never exceed 3 garments—the inference engine caps layering at three.
• Merchandise capsule drops as pre-mapped trios (e.g., 'Studio Sage Set') so the 'Complete the Look' App Block always has a full outfit to render on PDPs.
• Place the App Block on PDPs and the zero-CLS Floating Pill on collection pages via the OS 2.0 theme app extension; use seed-try-on:complete-look-add to fire /cart/add.js with items[] for all three SKUs.

{
  "custom": {
    "pair_with": [
      "gid://shopify/ProductVariant/44123456782410",
      "gid://shopify/ProductVariant/44123456782411",
      "gid://shopify/ProductVariant/44123456782412"
    ]
  }
}

Make Gym Selfies Disappear: Zero-Photo Storage by Default

Activewear shoppers upload the most sensitive photos in e-commerce—sports bras, leggings, post-workout sweat. Style Swap AI runs privacyZeroPhotoStorage: true out of the box: EXIF and GPS metadata are stripped in-browser, the image is compressed to a 1400px max dimension, transferred over TLS 1.3 and inferred in volatile RAM only. Nothing is ever written to your Shopify Admin, merchant database or CDN.

• Confirm privacyZeroPhotoStorage: true in the App Block settings—no persistent selfie storage and no shopper photos ever visible to your team in Shopify Admin.
• Client-side Canvas API compression (maxDim 1400px) plus EXIF/GPS stripping cuts upload weight on mobile gym Wi-Fi and eliminates geolocation metadata.
• WeVisOne-VL Reasoning Segmentation locks hair, face and hijab mattes—edges stay clean on modest activewear like longline tops and sports hijabs.
• Merchandise the privacy promise: add a 'Zero Selfie Storage' trust badge beside the try-on CTA and mirror it in your privacy policy—decisive for UAE PDPL, Saudi PDPL and luxury activewear shoppers.

5 Lighting Remixes + 9:16 Story Cards Built for Studio-to-Street

Turn one try-on into a scroll-stopping 1080x1920 Story card. Match the atmosphere to the use case—Studio Editorial for gym-floor sets, Golden Hour for run-club drops, Desert Resort for GCC modest athleisure capsules—then let shoppers share in one tap via the Web Share API. The native HTML5 canvas compositor renders client-side, so sharing never breaks zero-photo storage.

• Map remixes to merchandising moments: Studio Editorial (gym/studio), Golden Hour (outdoor training), Evening Lounge (recovery layers), Parisian Street (studio-to-street), Desert Resort (resort & modest capsules).
• Generate 9:16 Story cards at 1080x1920 entirely in-browser—no server round-trip—so WhatsApp and Instagram Story shares stay privacy-compliant.
• Track share intent with the seed-try-on:open and seed-try-on:complete-look-add DOM events; pipe them into GA4 or your analytics layer as micro-conversions.
• Use Quick Mode 1K (~8–10s) for on-PDP previews; reserve Best/Studio Mode 2K for retargeting ads, lookbook heroes and launch-day Story creative.

Closed-Loop Attribution: 'Tried On Product' Events + Paid-Order Webhooks

Connect Style Swap AI to Klaviyo for server-side 'Tried On Product' events and reconcile them against Shopify orders/paid webhooks to attribute revenue, AOV and bundle conversion down to the SKU. Activewear brands get a deterministic answer to 'did the 3-piece try-on actually sell the set?' instead of guesswork.

• Klaviyo sync unlocks on the Runway plan ($49.99/mo) and Unlimited Fashion plan ($199.99/mo); event payloads carry product/variant IDs, try-on mode and bundled SKUs.
• Reconcile try-on events with orders/paid webhooks to compute bundle-vs-single AOV lift per collection—e.g., does the Sage trio outsell the standalone legging?
• Build Klaviyo flows: 'abandoned look' (seed-try-on:open without add-to-cart within 24h), post-try-on cross-sell featuring the paired leggings/jacket, and win-back on restocked sizes.
• On headless Hydrogen/Next.js storefronts, pass try-on source as a UTM parameter or custom event property through the signed App Proxy at /apps/try-on/*.