Beyond the banner: The power of Web AI to personalize paid rich media ads
summary
Geenee demonstrates real-time, client-side web AR ads that served 20 million people in 30 days and raise apparel clickthrough from ~0.4% to ~0.7% (about 75% better) with a median engagement of 31 second median time, supporting try‑ons for handbags, hats, watches and soft goods in the ad unit itself. Inference dominates frames (75–85% of per-frame time), so performance gains from WebGPU and WebNN and compact models directly enable richer rendering and cloth simulation. The stack runs in the browser via virtual try-on ads and keeps camera data on-device.
AI-written summary · about
chapters (11)
- 0:00Web AI for personalized ads
- 2:09Live demo: Web AI AR ads
- 3:19Generative 3D models in ads
- 4:22Performance metrics for web AR
- 5:28Real-world applications beyond mobile
- 6:31Technical foundations of web AR
- 9:05Web AR building blocks
- 10:30Edge computing and latency in web AR
- 11:12Browsers as OS
- 12:39Performance with WebGPU and WebNN
- 13:24Client-side AI for privacy
related talks
24mWeb Apps of the Future With Web AI
45mPrivacy-First In-Browser Generative AI Web Apps
23mAR the next great digital ad format
16mThe Web is AI Ready—maximize your AI web development with WebNN
21mUnlocking Client-Side AI in Your Web App
39mThe AI PC and the Era of the Modern Web
29mClient side Web AI Agents Building smarter user experiences for a future age
32mPrivacy-first in-browser Generative AI web apps