Writing & MediaFree
CMS Migration Redirect and Asset Audit
Send your content export and new web address pattern. Get a redirect map, a check of every embedded file, and the outside-linked pieces to protect first.
Most CMS migration checklists center on a redirect map: every old page address paired with a new one and a permanent redirect, prioritized by traffic or outside links. Google's own site-move guide recommends exactly that mapping, and adds that embedded images, videos and other files need moving too. The common one-row-per-page templates have nowhere to track those files, and no column for the pieces other sites cite, which are the ones a broken migration hurts most.
On The Bellwood Courier's migration, an invented paper moving 4,180 archived URLs, 612 were low-value tag and search pages retired rather than redirected, leaving 3,568 real pieces to map. Those pieces carried 10,185 embedded asset references across 6,240 unique files, and 5,120 of those were self-hosted on the old CMS's own domain rather than an outside service. Checked against the new system's asset manifest, 433 had never been imported at all, leaving 501 pieces, 14 percent of the archive, with a broken image the day the redirects go live.
The same pass crosses the 3,568 real pieces against a backlink export. 287 carry at least one link from another site, 1,340 links in total, and 24 of those 287 still had no new address mapped, while lower-traffic pages further down the list already had one. Run this before the redirects go live, so the pieces readers, other publications and researchers depend on are not the ones found broken after launch. To check what a story claimed before it moves, the fact-check pack keeps each piece's sign-off record.
A redirect map is checked at the page. A reader hits the asset one level down
Every migration guide agrees on the page-level convention: old address, new address, redirect type, priority by traffic or backlinks, a test-status column. Published redirect map templates mention images and PDFs, if at all, as an aside inside the same row structure built for pages. Nothing in that structure catches a file that moved on its own schedule, because an image's address is not the article's address, and a redirect written for the page never looks at it.
On the worked archive, the 6,240 unique embedded files break down by hosting location first. 5,120 are self-hosted on the old CMS's own asset domain, which breaks the moment that domain goes dark; the other 1,120 are hosted by an outside wire service or embed provider, which survives untouched. Only the self-hosted set is at risk, and checking it against the new system's own import manifest, rather than assuming a successful page migration carried its media with it, is what finds the 433 that never made the trip.
The citation cross-check runs the same logic one level out. A backlink or citation export names which old URLs something outside the newsroom, another publication, an academic paper, a court filing, actually points at. Crossing that list against the redirect map's own completion status surfaces the pieces where an outside citation is about to break, 24 of them on the worked archive. That list is not the same list a traffic report would produce.
How it works
Send the export
Attach the content export with its addresses and article bodies, and describe the new address pattern it is moving to.
Split and map
River separates real content from low-value pages, then proposes a target URL for each piece that is actually moving.
Check assets and citations
Every embedded file is checked against the new system's import log, and every page against your backlink export.
Read the sequence
A Doc sets the order to fix things; a Sheet holds every row for whoever sets up the redirects.
What you get
- Splits the export into real stories and low-value listing pages, so retired tag and search pages skip redirects
- Traces every embedded image, PDF and video to one file, so a reused photo is counted once
- Flags every old-server file missing from the new system's import log, before readers find it
- Names every piece carrying a missing file, since a page that redirects cleanly can still be broken inside
- Crosses the archive against a backlink export to rank pieces by how much other sites link to them
- Flags the most-linked pieces still missing a new address, so they outrank pages nobody links to
- Puts the redirect map and file check in one sortable Sheet, plus a Doc setting the work order
Common questions
Isn't a redirect map enough on its own?
It is enough for the page itself, which is what most migration checklists verify. It is not enough for what the page embeds. An image, PDF or video carries its own address, and a redirect written for the page never touches the file inside it. On the worked archive, 14 percent of migrated pieces still showed a broken image once the file check ran.
What if we don't have a backlink or citation export?
The redirect map and the file check both run without one; only the outside-link ranking needs it. Search Console's Links report lists your most-linked pages and can be exported, and backlink tools offer similar exports. Without one, River ranks by whatever traffic figures your export includes and says so, and you can rerun the ranking when the backlink data arrives.
Why do low-value pages get retired instead of redirected?
Because a tag, author or search-results page usually has no single equivalent in a new CMS, and forcing a one-to-one redirect sends visitors somewhere that does not answer what brought them there. The worked archive retired 612 of 4,180 addresses this way, about 15 percent, pointing them at the nearest section page or letting them end, rather than mapping each to a guess.
What counts as a missing file?
A file embedded in a piece that is moving, stored on the old CMS's own server rather than an outside service, with no matching entry in the new system's import log. It is a direct comparison against the log the new system produced, not an assumption that a successful page move carried its images along. Without a log, River lists the files at risk and says they are unverified.
Does this replace testing the redirects after launch?
No. Google's site-move guidance recommends monitoring traffic on both the old and new site after the move and keeping redirects for as long as possible, generally at least a year. This audit runs before that, so the list you test against already includes the files and outside links a page-only map would miss.
Where does this sit in the rest of the desk's work?
Before anyone writes the redirects, as a one-off project worth its own slot on the editorial calendar. The Sheet goes to whoever sets up the redirects, engineering or the CMS vendor; the Doc goes to the editor deciding what is retired. If a contributor's commission agreement limits how long a piece stays online, check it before carrying that piece over.
Built by
Chandler Supple
Co-founder & CEO, River
Chandler has spent the last six years working in AI, building tools for millions of users and Fortune 100 companies. In 2025, he co-founded River to build a new kind of operating system for AI agents.
CMS Migration Redirect and Asset Audit
Describe what you need, add any files you have, and your workspace opens with the work already underway.