Why does AI cite the wrong version of my page?
Because the same page is reachable at more than one address, such as with tracking parameters, with and without www, or over HTTP and HTTPS, and engines may pick any of them. Choose one preferred URL and point the others at it with permanent redirects or canonical tags. Google describes both methods and says redirects are the strongest signal.
Free to read, no signup. Nova finds duplicate versions of your key pages.
What Google says about duplicate URLs
Straight from Google's documentation. Other engines have not published the same guidance.
Why duplicates appear
Google says the same content can be reachable at several addresses, commonly through tracking parameters, host variants such as www or not, HTTP versus HTTPS, and alternate paths.
Consolidate duplicate URLsThree ways to state a preference
Google ranks permanent redirects as the strongest signal, rel=canonical as strong, and a sitemap as the weakest. They can be combined, and if you state no preference Google chooses a version itself.
Consolidate duplicate URLsFive steps to one preferred version
Do it by hand with these steps, or hand the whole thing to Nova.
Find the version being cited
Ask the engines your key questions and note the exact URL each cites, including any parameters or old paths.
Nova: lists the URLs engines cite for your prompts.
Find every version that exists
Test www and non-www, HTTP and HTTPS, trailing slashes, tracking parameters and old paths for each key page.
Nova: finds the duplicate versions of your key pages.
Choose one preferred URL
Pick the clean, current address you want everyone to use, and use it in links, sitemaps and profiles.
Nova: suggests the preferred URL for each page.
Redirect or canonicalise the rest
Use a permanent redirect when a version should be retired, and rel=canonical when it must stay reachable. List the preferred URL in your sitemap.
Nova: drafts the redirect and canonical rules for approval.
Check again later
Re-ask the engines in a few weeks and check which URL they cite. Allow time for recrawling.
Nova: re-checks the cited URLs and reports the change.
Short on time? Nova, your AI brand visibility expert, runs this Maneuver on your brand and tracks the result.
Take it to NovaDuplicate versions and the fix
| Cited URL | Preferred URL | Issue | Fix |
|---|---|---|---|
| http://www.[domain]/pricing | https://[domain]/pricing | HTTP and www | Permanent redirect |
| [domain]/pricing?utm=[source] | https://[domain]/pricing | Tracking parameter | Canonical tag |
| [domain]/old-pricing | https://[domain]/pricing | Retired page | Permanent redirect |
Nova then re-checks which URLs engines cite, so you see whether they now use the preferred one.
Take this Maneuver into your own AI agent
Get duplicate-page-fix.md, a ready-to-use skill file with a version-hunting routine, a preferred-URL table and a re-check routine. Drop it into Claude, Cursor or any agent that reads SKILL.md files.
Skill file for Claude, Cursor and other agents
- A duplicate URL hunting routine
- A preferred-URL table
- A redirect versus canonical guide
Questions, answered.
Relevant reads
All articlesLet Nova find your duplicate pages
Nova opens with this Maneuver loaded. You approve every change before it ships.