THE PROBLEM
Customer content team is publishing 10–20 blog posts per week as part of a large-scale SEO initiative. Their workflow involves pulling images from stock sources (e.g. Getty), uploading directly to the DAM, and passing them to development. They are not designers, and requiring them to manually resize or optimize images in tools like Figma before upload creates friction and slows publishing velocity. The current workaround is to have UX designers optimize every image before handoff — which has become a recognized bottleneck.
Beyond workflow, there is a real page performance concern. Technical contact confirmed that current page load times exceed 2 seconds, with a target of under 1 second.

CURRENT STATE
Customer uses WebP format on their website, but the WebP files are not actually optimized — a 10MB source image produces a ~2.4MB WebP that is served at the same size across all devices (desktop, tablet, mobile).
All 20,000–50,000+ assets flow through CMP/DAM (using CloudFront as the CDN).
The CMS side uses Cloudflare, which has a plugin called Polish that performs adaptive image rendering — it detects the request origin (device type / breakpoint) and serves an appropriately sized file (e.g. ~10KB for mobile, ~1MB for desktop from the same source URL).
CMP/CloudFront has no equivalent capability today. This is the gap.

WHAT CUSTOMER IS ASKING FOR
An adaptive image rendering capability native to CMP, similar in function to Cloudflare Polish. Specifically:

When an asset is served from the DAM/CMP, the system should detect the request context (device type / breakpoint) and deliver an appropriately optimized version of the image.
This should be automatic — no manual re-upload, resizing, or format conversion required from the content team.
The ask is not to edit the image (change aspect ratio, crop, etc.) but to render it intelligently based on context.
Ideally, this would work for all asset types coming out of CMP, not just blog images.