Content Decay: How to Prioritise Refreshes Without False Positives

The true cost of content decay is often the wrong refresh

Content maintenance costs more than the writing itself. A meaningful refresh can require research, editing, subject-matter review, design or development changes, quality assurance and then measurement. The opportunity cost matters too: every hour spent rewriting a page that did not need it is an hour not spent on a page with clearer business value.

The opposite mistake is also expensive. A page can become less useful because facts have changed, the searcher's task has changed, the offer is obsolete, competing pages on your own site now overlap, or the page no longer helps users take the next step. Leaving that page untouched can mean fewer qualified visits, weaker conversions, avoidable support questions or misleading information.

The useful question is therefore not “How old is this page?” It is “What evidence shows that this page is no longer doing its job, and is changing the content likely to solve the problem?” Google’s guidance on debugging Search traffic drops starts with diagnosis: compare periods, look for patterns, and isolate the affected pages, queries, devices, countries or search appearances before deciding on a fix.

What content decay is — and what it is not

Content decay is an industry shorthand, not a Google classification. In this guide, it means a page has become materially less fit for its intended job because its information, relevance, usefulness or business role has deteriorated.

That is different from a general performance decline. A page can lose clicks while its content remains accurate. It can have poor Largest Contentful Paint (LCP) while the article itself remains useful. It can have a low engagement rate because visitors found a phone number immediately. It can also be several years old and still answer the query perfectly.

Google’s people-first content guidance explicitly warns against changing dates to make pages look fresh when the content has not substantially changed, and against removing or adding large amounts of content simply to make a site seem fresh. It also says there is no preferred word count. See Google’s helpful, reliable, people-first content guidance.

That distinction prevents a common false positive: treating age, a single metric or an audit warning as proof that a rewrite is required.

Capture a baseline before changing the page

Before editing, record enough evidence to explain why the refresh is happening and to judge whether it worked. For a commercially important page, capture:

• Page identity: URL, current title/H1, publication date and last meaningful update where available.
• Search performance: clicks, impressions, click-through rate (CTR), the queries that surface the page, and how those metrics changed against a sensible comparison period.
• On-site outcomes: organic sessions, engagement rate, key events such as enquiries or purchases, and any page-specific conversion measure you actually trust.
• Content evidence: claims that are now wrong or incomplete, dead or weak sources, changed product details, changed regulations or standards, and questions the page no longer answers.
• Internal context: pages that link to it, pages it links to, and other ScanMySEO content targeting substantially the same intent.
• Refresh cost: expected research, editorial, expert-review, implementation and QA time.

Google explains that Search Console measures what happens before the visit—impressions, clicks and queries—while Google Analytics measures what people do on the site. Using both gives a more complete picture, but the numbers are calculated differently and should not be expected to match exactly. Google’s Search Console and Analytics guide is useful when building this baseline.

If generative-AI visibility matters to your business, record that too rather than guessing. Google announced dedicated Generative AI performance reporting in Search Console in June 2026 and said the rollout reached all websites worldwide on 31 August 2026. Treat it as an additional visibility signal, not as proof that a page needs rewriting. See Google’s Generative AI performance report announcement.

False-positive checklist: diagnose before you refresh

Use the following sequence on the page that appears to be declining. The order matters because it prevents you from rewriting content to solve a problem that sits somewhere else.

1. Confirm that the decline is real. Compare like-for-like periods. Check year-over-year patterns where seasonality is plausible. A short dip, holiday effect or temporary demand change is not enough evidence for a rewrite.
2. Isolate the affected page and query set. In Search Console, compare the periods and inspect both Pages and Queries. If the entire site or a large template group moved together, investigate a broader technical, indexing or market change before blaming one article.
3. Separate visibility from click-through problems. If impressions are broadly stable but clicks fall, inspect CTR, the queries involved and how the result is presented. Google notes that a clicks-only decline can indicate a title or snippet problem rather than a loss of underlying search visibility.
4. Check whether search intent has actually changed. Read the queries that still surface the page and compare them with the task your page solves. If the page explains a concept but searchers now need diagnosis, comparison or implementation help, the refresh may need structural change rather than extra paragraphs. For a deeper workflow, see ScanMySEO’s search intent mismatch guide.
5. Re-fact-check the page. Look for obsolete prices, screenshots, feature names, legal references, platform interfaces, version-specific instructions, broken citations and statements that have become more nuanced. One important outdated claim can justify a targeted refresh even if traffic is stable.
6. Check on-site behaviour without turning it into a ranking signal. A fall in engagement or key events can show that the page is less useful to your visitors or less commercially effective. It does not, by itself, prove that Google sees the content as low quality. Investigate the landing page, audience, traffic source and measurement setup before drawing conclusions.
7. Separate technical problems from editorial problems. Slow loading, broken rendering, accidental noindex, canonical mistakes, redirect problems or template changes can reduce performance while the words remain perfectly good. Fix the technical cause first when that is the actual failure.

How common “decay” signals become false positives

Google’s current Search Console documentation lets you compare pages and queries and sort by change between periods. It also advises focusing more on trends in impressions and clicks than on absolute average position. The Performance report task guide is a better starting point than a blanket “rewrite anything that dropped” rule.

Prioritisation framework: evidence first, business value second

Once you have diagnosed the cause, decide priority using two questions:

1. How strong is the evidence that the content itself has decayed?
2. How valuable is this page to the business or user journey?

Business value is broader than keyword volume. A page may deserve priority because it generates enquiries, supports a sales conversation, answers a costly support question, earns useful links, helps existing customers or is required for trust. Conversely, a high-impression page with weak business relevance may not justify a major rewrite.

Use a qualitative rule rather than a fake precision score: priority rises when evidence of decay, business value and confidence that content is the cause are all high.

Choose the smallest action that solves the diagnosed problem

A “refresh” does not have to mean rewriting the whole page.

• Keep: Use when the page remains accurate, useful and aligned with intent. Correct genuine errors if you find them, but do not rewrite for the sake of activity.
• Targeted refresh: Update the specific decayed element—such as an obsolete feature name, outdated example, missing caveat, weak source or changed user question—while preserving sections that still work.
• Structural refresh: Use when intent has changed or the page answers the right topic in the wrong format. Reorder the page around the current task: understand, diagnose, decide, act and verify.
• Consolidate: Use when two pages compete to solve substantially the same task and one stronger page can serve users better. Decide which URL should remain before moving material.
• Retire: Use only when the page no longer has a useful purpose and there is no reason to preserve it. If equivalent content has moved, use an appropriate permanent redirect. If the content is genuinely gone and has no replacement, Google recommends returning a real 404 or 410 rather than a misleading “soft 404”; see its crawling error guidance.

Do not delete pages merely because they have low traffic. Low traffic can be a valid outcome for a narrow but necessary page, and removing a URL can affect internal links, external links, customer journeys and historical visibility.

Do not create “AI refresh” busywork

Google’s May 2026 guidance for generative-AI features reinforces the same principle: normal SEO fundamentals still apply. Google says there is no ideal page length for AI Search, no requirement to split content into tiny “AI-friendly” chunks, no special schema needed for generative AI, and no need to rewrite pages in a special style just for AI systems. Its emphasis is on useful, unique, non-commodity content. See Google’s generative AI optimisation guide.

That means a refresh should add missing value, correct genuine decay or reduce reader effort—not chase a target word count, add generic FAQs, insert extra headings, or reformat a good page simply because AI search exists.

Use audit tools as evidence, not verdicts

Automated audits are useful for surfacing patterns quickly, but a flag is a prompt to investigate, not proof that a page needs a rewrite. For ScanMySEO, that means treating findings and severity as prioritisation aids rather than as Google scores or guarantees.

Google’s 2026 guidance on third-party SEO tools and advice makes the boundary explicit: third-party tools do not have access to Google’s internal ranking data and cannot guarantee performance. Use audit data alongside Search Console, Analytics, direct page review and business context.

Measure whether the refresh worked

Write down the hypothesis before publishing. Examples:

• “Clicks fell while impressions were stable; we expect a clearer title and description to improve CTR for the same query set.”
• “The page still ranks for the right topic, but the instructions describe a retired product interface; we expect the update to reduce confusion and support requests.”
• “The page attracts the right organic visitors but few complete the enquiry form; we expect a clearer decision path to improve the relevant key event.”

After the change, compare the same metrics you captured before it. Allow enough time for the page to be crawled, for demand to normalise and for your sample to become meaningful. Do not claim success because one metric moved for a few days.

Also record what changed. A useful refresh log can be as simple as: date, reason, sections changed, sources replaced, hypothesis and the metrics you plan to review. That makes the next refresh easier to diagnose and prevents teams from repeatedly “optimising” the same page without knowing what previous work achieved.

A practical decision rule

Refresh when you can point to a real problem, show that the page matters, and explain why changing the content is a plausible fix. Keep the page when the evidence is weak. Fix technical problems as technical problems. Treat seasonality as seasonality. Consolidate overlap deliberately. Retire content only after checking what would be lost.

The most expensive content-refresh programme is not the one with the most old pages. It is the one that spends scarce editorial and technical time without first proving what needs to change.