CONTENT WRITING

Refresh a decaying article

Use when an old post is losing rankings or traffic and you need a specific update plan rather than a rewrite.

refresh-decaying-article.md

Download .md

You are auditing an existing article that has lost traffic, and producing an update plan. Do not rewrite the whole thing.

Article: {{ARTICLE_TEXT}}
Published: {{PUBLISH_DATE}}. Last updated: {{LAST_UPDATED}}
Traffic and ranking change: {{PERFORMANCE_DATA}}
What has changed in this topic since publication, as far as I know: {{KNOWN_CHANGES}}
Queries it currently ranks for: {{QUERY_LIST}}

Output:
1. Decay diagnosis table: Possible cause (stale facts, intent shift, thin section, lost coverage of a query, dated examples, competitor depth) | Evidence in the article or data | Confidence (high/medium/low).
2. Section-by-section verdict table: Section | Keep as is | Update | Rewrite | Cut | What specifically to do.
3. New sections needed, each with the query or reader need that justifies it.
4. Facts and figures that must be re-checked before republishing, quoted from the article with their apparent age.
5. What to leave alone, and why touching it would risk the rankings it still has.

Rules:
- Do not recommend rewriting sections that are still accurate and still ranking.
- Do not supply replacement statistics from memory. Mark them as to be sourced.
- Rank the section verdicts so the highest-impact change is first.

Fill in before running

Replace each placeholder with your own detail. The more specific you are, the less the model invents.

{{ARTICLE_TEXT}}
{{PUBLISH_DATE}}
{{LAST_UPDATED}}
{{PERFORMANCE_DATA}}
{{KNOWN_CHANGES}}
{{QUERY_LIST}}

Getting a better result

Feed it the query list from Search Console; intent shift shows up there before it shows up in traffic.
Section 5 is the important one - most refreshes lose rankings by rewriting what worked.
Re-run after the update with the new text to check nothing was dropped.

Questions about this prompt

When should I audit the article rather than rewrite it?

When it is losing traffic but still has something working. If the piece is wrong end to end, rewriting is cheaper than auditing it. This prompt is explicitly told not to rewrite the whole thing, so its value sits in deciding which sections to leave completely alone.

Which inputs does the decay diagnosis need?

The article text, publish and last updated dates, the traffic and ranking change, what you know has changed in the topic, and the query list from Search Console. The query list earns its place: intent shift shows up there before it shows up in the traffic graph.

What is in the update plan?

A decay diagnosis table with confidence levels, a section by section verdict, new sections each justified by a query or reader need, figures to re-check with their apparent age, and what to leave alone. Section 5 is the important one. Most refreshes lose rankings by rewriting the part that still worked.

What turns a refresh into a new problem?

Filling in the replacement statistics yourself from memory. They are marked to be sourced for a reason, and guessing turns a refresh into a new accuracy problem carrying a fresh date. Re-run the prompt on the updated text afterwards to confirm nothing that was ranking got dropped in the edit.