NovAn old article should not compete forever with the new canonical guide
How to decide whether an older article should be merged, redirected, narrowed or retained after a stronger guide takes over the main user task.
This article reflects the named expert’s practical perspective. See NovAsia’s editorial policy for how material is prepared and reviewed.
Content rarely arrives in the order a clean information architecture would prefer. A site publishes a short answer, then a comparison, then a longer guide. Two years later, the editorial team finally creates the page that should probably have existed from the start. The new guide is broader, better sourced and more useful, yet the old article still has links, impressions, bookmarks and a perfectly valid URL.
That does not automatically make the older page a problem. The problem begins when nobody can explain what job it has now.
I use “canonical guide” here in the editorial sense first: the page the site wants people to treat as the main answer. Only after that decision is clear should the team reach for technical canonicalisation, redirects or other indexing controls.
Two pages can mention the same subject without competing
An old article and a new guide may share terminology but serve different stages of the decision.
A broad guide on foreign property ownership could map the available structures and the questions a buyer needs to resolve. A narrower article could focus on one specific document or one condition that becomes relevant later. Those pages overlap in subject matter but not in purpose. They can reinforce each other if the boundaries are explicit.
The harder case is when both pages answer the same question with the same mechanism and essentially the same conclusion. The old article may be shorter and the new guide more polished, but the user still sees two versions of one answer.
A keyword comparison will not diagnose that reliably. I compare the pages by three things: the primary user question, the evidence each page uniquely contains, and the next action a reader is prepared to take. If those are almost identical, the site probably has a role conflict, not a vocabulary problem.
Consolidation starts with content, not with a tag
When the new guide genuinely replaces the old answer, I would first recover anything still valuable from the older article. A specific example, a useful table, a precise explanation or a historical note may belong in the new page. Consolidation should not destroy good information merely to simplify URLs.
Once the main answer is complete, the old URL can be handled according to its remaining purpose. If it no longer has one and there is a clear replacement, a permanent redirect may be a sensible outcome. A user following an old link reaches the current answer instead of landing on an abandoned page.
If the old URL is simply a technical or near-identical variant of the same content, canonicalisation may be relevant. Google describes canonicalisation as choosing a representative URL from duplicate or very similar pages. Yandex describes a similar grouping of identical or similar content. Both also make an important point: a declared canonical is a preference or recommendation, not an absolute command.
That is why I would not use `rel="canonical"` as a generic label meaning “we prefer the newer article”. If the two pages are materially different, the tag may be describing the relationship badly.
Keeping the old article can be the better decision
Sometimes the new guide reveals that the old page was useful, just badly positioned.
Perhaps the old article has a strong narrow question buried under an overly broad title. Instead of deleting it, the team can tighten its scope. Remove the parts that duplicate the main guide, deepen the specific issue and link the two pages in a deliberate hierarchy.
Another valid role is historical context. A page about a market condition or rule at a particular date can remain useful if it is clearly labelled as historical and does not masquerade as current guidance. The new guide then handles the current decision while the older page preserves a dated record.
The key is intentionality. “It still gets traffic” is not enough to define a page's role. Traffic can be evidence that people continue to need something from the page, but the team still has to identify what that thing is.
Internal links reveal whether the decision is real
A content decision is incomplete when the rest of the site continues to behave as though both pages are primary.
I review links from hubs, project pages, older articles, breadcrumb paths, related-content blocks and any navigation that repeatedly points to the old URL. If the new guide is now the main answer, the architecture should say so consistently. If the old article retains a narrow role, anchor text and surrounding context should make that role understandable.
The XML sitemap should also reflect the intended structure, but a sitemap is not a substitute for it. Google explicitly states that a sitemap can help discovery but does not guarantee crawling or indexing. Keeping two competing URLs in a sitemap does not teach the engine why one is the main guide.
Search data helps diagnose the overlap, but it does not make the editorial decision
Search Console can show whether both URLs receive impressions for similar queries and which page Google considers canonical in relevant cases. That is useful evidence. It does not tell the editor what the site should mean.
A page may lose impressions after a redirect or consolidation and that can still be the correct outcome if the stronger guide now answers those searches. Conversely, a new guide may gain visibility while the site accidentally discards a useful specialist article that deserved its own role.
Therefore, I would judge the result by a combination of search behaviour and user architecture. Are broad queries going to the broad page? Does the narrow article attract a narrower need? Do internal links reinforce that distinction? Can a user understand why both pages exist?
Every legacy URL should end with a stated reason to exist
Mature sites accumulate history. Trying to remove every old page is as crude as keeping every old page forever.
The healthier standard is accountability. The team should be able to put each legacy article into a category: replaced and redirected, merged into the main guide, retained as a narrower answer, preserved as historical material, or still unresolved. The unresolved group should shrink over time.
A new flagship guide does not become canonical because the editorial team calls it definitive. It earns that role by becoming the clearest main answer and by being supported by the surrounding architecture. The older article then deserves either a new, distinct job or a clean retirement. Endless competition is not a strategy.
Sources
- Google Search Central — “What is canonicalization?”: explains representative canonical URLs for duplicate or very similar pages and notes that preferred canonical signals are not absolute rules. Accessed 2026-10-06.
- Google Search Central — “Fix canonicalization issues”: recommends checking Google's selected canonical and considering whether pages are sufficiently different. Accessed 2026-10-06.
- Yandex Webmaster — “Canonical URLs”: describes canonical grouping for identical or similar pages and treats `rel="canonical"` as a recommendation that may be ignored in some cases. Accessed 2026-10-06.