Should you merge two cannibalizing pages or let them coexist?

Why does Google differentiate cannibalization based on current performance?

Keyword cannibalization is often presented as a scourge to be systematically eradicated. However, John Mueller seriously nuances this approach. When two pages compete for the same semantic territory without managing to establish themselves in the results, scattering signals (backlinks, anchors, CTR, dwell time) weakens the overall visibility of the site.

In this specific case, consolidating via canonical tag allows for channeling all these signals to a single reference URL, thus increasing its chances of ranking solidly. It’s a principle of concentrating forces: instead of two weak pages, we bet on a robust page.

When is it still relevant to maintain two distinct pages?

The logic completely reverses when both pages are already positioned in the top 1 or 2. Merging would mean sacrificing a dominant position without a tangible gain. Google acknowledges here that certain queries carry semantic or intentional nuances distinct enough to justify two results from the same domain.

In practical terms? If you rank #1 for "SEO agency Paris" and #2 for "SEO consultant Paris," maintaining two dedicated pages allows you to lock down the SERP and capture clicks on slightly different intents (team need vs individual expert). Merging would destroy this dual presence.

How can you assess if two pages are truly cannibalizing each other?

Cannibalization is not binary. It depends on the volatility of positions, the frequency of oscillation between the two URLs in the SERPs, and their combined CTR. If one rises when the other falls cyclically, it's a strong signal of internal competition.

Analyze the Search Console: compare impressions, clicks, and average positions of the two pages over 3 months. If neither consistently exceeds position 5-6 and their curves intertwine, merging may unleash potential. Conversely, two stable pages in the top 3 with distinct main queries are not cannibalizing — they coexist.

• Destructive Cannibalization: two unstable pages in positions 8-15, neither gaining lasting traction
• Productive Coexistence: two pages in the top 3 on complementary semantic variations
• Decisive Criterion: stability of positions and distinction of main queries in the Search Console
• Recommended Action: test canonical for 4-6 weeks while monitoring changes in organic traffic and positions
• Trap to Avoid: merging dogmatically without analyzing the actual performance of the two URLs

Is this recommendation consistent with field observations?

Yes, and it aligns with a well-known empirical rule among seasoned SEOs: cannibalization is only a problem if it creates instability. In practice, we regularly observe sites that dominate a SERP with two URLs in positions 1-2 — often because the pages cater to different micro-intents (comparison vs buying guide, tutorial vs definition).

However, Mueller’s recommendation remains vague on the exact threshold. What do we mean by "struggling to rank"? Position 5? 10? 20? And over what period should we measure this difficulty? [To be verified]: no quantitative data has been provided by Google to draw a clear boundary between "beneficial merging" and "optimal maintenance."

What risks does this consolidation strategy hide?

Merging two pages via canonical means betting that the canonical URL will absorb all signals from the deprecated page. But if the latter historically captured quality backlinks or ranks on specific long-tail variations, the redirection may cause a temporary loss of traffic while Google recalculates the consolidated authority.

Another pitfall: if the two pages address partially different subjects, merging risks diluting the thematic relevance. A typical example: "best CRM" and "CRM comparison 2025" may seem redundant, but if the first targets a decision-making intent and the second an informational intent, merging them can confuse the signals sent to Google about the real intent of the page.

In what cases does this rule not apply?

This binary logic (merge if weak, maintain if strong) overlooks low domain authority sites. For a new site, even two pages in positions 8-12 may be worth keeping separate if they each generate qualified traffic on low-volume keywords. Merging won’t change anything if the underlying problem is a deficit of backlinks or overall authority.

Similarly, this approach does not account for seasonal or event-based contexts. Two pages may cannibalize each other temporarily during a search peak, then coexist peacefully for the rest of the year. Hasty decisions = strategic errors.

How to concretely decide to merge or maintain two pages?

Start by extracting Search Console data for both pages over 6 months: impressions, clicks, average CTR, average position. If neither consistently exceeds position 5 and their curves overlap without ever stabilizing, it's a clear signal of destructive competition.

Next, map the main queries of each URL. If they target exactly the same keywords with identical volumes, merging via canonical is relevant. If each page captures distinct variations (plural/singular, with/without location, question/statement), maintaining two URLs can maximize coverage.

What mistakes to avoid during consolidation?

Never merge without having audit the backlinks of the deprecated page. If it concentrates quality links pointing to specific anchors, these signals must be preserved via 301 redirects to the canonical page — or better, by keeping the page and optimizing it differently.

Another classic pitfall: choosing the wrong URL as canonical. Always prioritize the one with the oldest history, the best backlink profile, and the most concise URL. If you get it wrong, Google may ignore your directive and choose the canonical version itself, creating even more confusion.

How to check if the consolidation is yielding results?

Deploy the canonical tag and monitor weekly the evolution of positions and organic traffic over 4 to 6 weeks. A good sign: the canonical page gradually climbs and stabilizes its position, while the deprecated page disappears from the SERPs without any loss in overall traffic.

If, conversely, traffic drops or the canonical page stagnates, it means that the two pages were likely capturing different intents. In this case, remove the canonical and work on the editorial angle of each page to better differentiate them.

• Extract 6 months of Search Console data for the two competing URLs
• Map the main queries and identify exact overlaps
• Audit backlinks and anchors for each page using Ahrefs or Majestic
• Choose as canonical the oldest, best-linked, and most concise URL
• Deploy the canonical tag and monitor positions + traffic over 6 weeks
• If traffic drops > 10%, remove the canonical and rework editorial angles