Triage -What “underperforming” really means in 2026

Brett Tabkeon

Phase 2: Triage

What “underperforming” really means in 2026

Phase 2 is where most content audits lose discipline. This is the point where teams fall back on familiar metrics and start defending pages that look successful in isolation. Rankings and sessions still matter, but they no longer explain visibility, suppression, or disappearance in modern search systems.

In 2026, underperformance is less about where a page ranks and more about how it behaves once surfaced. Pages can rank, earn impressions, and still actively fail to justify their existence when measured against engagement, clarity, and competition within the same site.

Modern Signals of Underperformance

These signals appear repeatedly across mature sites. No single signal is a verdict, but patterns are rarely accidental.

  • Zero-impression URLs: Pages indexed but no longer considered relevant for any active demand
  • Impressions with no clicks: Content that surfaces but is consistently ignored
  • AI-displaced pages: Queries now resolved entirely by summaries or answer blocks
  • Declining query sets: Rankings that persist while engagement collapses
  • Multiple pages declining together: A strong signal of intent overlap or internal competition

Triage replaces intuition with categorization. Pages are no longer judged as “good” or “bad”, but evaluated by the signals they emit. Some fail because they are obsolete. Others fail because they are redundant. Others fail because they are incomplete, unfocused, or overshadowed by stronger internal neighbors.

Fix

Pages showing demand signals but lacking clarity, depth, or alignment.

Merge

Multiple URLs competing for the same entity, question, or task.

Remove

Pages with no measurable demand that create crawl or relevance noise.

Triage Decision Flow (Signals → Outcome)

Use this as a fast first-pass. It is not a verdict, it is a disciplined way to sort URLs before deeper review.

Start: URL in your inventory

Has impressions (last 6–12 months)?

No
REMOVE
Zero demand signal

Yes
Has clicks?

Yes
FIX or KEEP
Protect demand, improve clarity

No
Intent overlap (same entity/task)?

Yes
MERGE
Collapse competing answers

No
FIX
Improve match and completion

Tip: do not use backlinks as a shortcut. Links influence the destination you choose, not the triage category.

The outcome of Phase 2 is not execution. It is reduction of ambiguity. Every URL should now sit in a provisional bucket, backed by observable behavior rather than opinion.

Phase 2 Deliverables

When triage is complete, every indexable URL should have a defensible preliminary outcome.

  • Pages flagged for improvement due to unmet demand or weak engagement
  • Groups marked for consolidation based on overlapping intent
  • Pages identified for removal due to zero demand or systemic redundancy
  • A documented rationale tied to measurable signals for each decision

Phase 1 reveals behavior without judgment. Phase 2 assigns meaning to that behavior, redefining what “underperforming” looks like when rankings and sessions are no longer enough.

Content Pruning Guide 2026:

As we publish this series, it will be a deep dive into content pruning we call the Era of Spray-and-Pray is over.


Here is what the content pruning series will cover:

  • Phase 1: Audit: How to audit without pre-existing bias
    We start with the mechanics of a modern content audit, using GSC, crawlers, and log data to identify pages that are quietly hurting site performance and not just dead weight content.

  • Phase 2: Triage: What "underperforming" really means in 2026 - When to fix, merge, or remove content.
    Rankings and sessions are no longer enough. We break down new signals like zero impression URLs, AI displaced content, and query sets that no longer produce clicks at all.

  • Phase 3: Consolidation: How to consolidate without losing authority
    We are going to cover redirects, internal link rewrites, canonical handling, and how to roll excessively thin posts into a single stronger resource without triggering ranking losses.

  • Phase 4: Slop on Top: How AI systems radically change the payoff
    Pruning is no longer just about rankings. We examine how cleaner content libraries improve citation likelihood, entity recognition, and visibility inside AI-generated answers. If the point isn't a click - ummm - what's the point again?

  • Phase 5: Measurement: How to measure success
    We close by redefining what "working" looks like, focusing on index health, impression quality, and how often your content becomes the source rather than the click.