Find pages competing for the same keywords, with overlap analysis, ranking impact assessment, and specific merge, redirect, or differentiation recommendations per conflict.
Share the client's site with Search Console connected. You get back a severity-ranked conflict list showing which page pairs are splitting authority, with a specific verdict per conflict: merge with a 301, sharpen each page's keyword focus, or archive the weaker one.
The team sees every instance where two or more pages compete for the same query. The analysis cross-references query-level data across all pages: where Google shows multiple pages from the same site, where position fluctuates because Google keeps switching which page to display, and where both pages rank but neither cracks the top 5, which is a classic sign of split authority.
For each conflict, both pages get scraped and compared: content scope, heading structure, keyword focus. Conflicts are grouped by severity. A high-volume keyword where both pages are stuck outside the top 5 is critical. One page ranking well while the other dilutes it is moderate. Minor overlap with little traffic impact is low. Each conflict comes with a specific verdict: merge page B into page A with a 301 redirect, keep both but sharpen each page's keyword focus so they stop competing, or archive the weaker page entirely.
Cannibalization does not show up in standard SEO audits. Two pages can both rank on page 1 for different queries and still cannibalize each other on a third query the team did not check. It also hides in position volatility: if a page fluctuates between position 5 and position 15 from week to week, it may be because Google keeps switching which page to show. The only reliable way to find it is to cross-reference query-level data across all pages, which is what this flow does with Search Console data.
The analysis found two pages that should be combined. This prompt produces the merged content: the best sections from each page, a consolidated keyword target, a redirect plan, and an updated title and meta description.
Merge the two cannibalizing pages into one. Take the strongest sections from each, consolidate the keyword targets, write a new title and meta description, and give me the 301 redirect instruction for the page being retired.
Sometimes two pages overlap on keywords but serve different intents. This prompt sharpens each page's focus so they stop competing: distinct keyword targets, differentiated titles, and internal linking to connect them.
These two pages should both exist but they're competing for the same keywords. Redefine the keyword focus for each one: give each a distinct primary keyword cluster, rewrite the titles and metas to differentiate, and recommend internal links between them so they support each other instead of competing.
A Juma Project is a shared space where the team stores everything Juma needs to know about a client. Create one project per client, add context as you go, and Juma will use what's relevant every time the team runs a flow. If the project already exists from other work, just add the items below.
Content Inventory
A list of all pages on the site with their primary keyword targets. This helps Juma spot overlaps faster: if two pages both target "email marketing best practices," that conflict is visible before even pulling GSC data.
Past Audit Results
Previous SEO audits that flagged cannibalization. Juma checks whether past conflicts were resolved and flags any new ones that appeared since the last review.
Site Architecture Notes
How the client's content is organized: categories, topic clusters, pillar pages. This context helps Juma distinguish intentional content overlap (a pillar page and its supporting posts) from unintentional cannibalization.
Add a short description in the project's info field that tells Juma what each file contains and when to use it. For example: "Content Inventory: full page list with keyword targets, last updated March 2026. Past Audit: January 2026 audit, check if past cannibalization issues were resolved. Site Architecture: topic cluster map showing pillar pages and supporting content."
