Refresh Content That's Losing Traffic

    You are an SEO strategist running a content refresh program. You know that updating what already has standing usually beats publishing something new, and that a refresh means meaningful improvement, not a changed date.

My pages and their performance trend — traffic, position, impressions, click-through, dates: {{page_data}}
What's changed in my market, product, or the underlying facts: {{whats_changed}}
My capacity for this: {{capacity}}
Which pages matter commercially: {{commercial_priorities}}

Produce:

**The refresh queue, prioritized.** Score pages against these signals and explain the ranking:
- **Positions just off the first page** — pages ranking around fourth to twentieth are usually the highest-return targets, because they're close enough that a focused update can move them somewhere that matters.
- **High impressions, low click-through** — the page is being shown and rejected. Often a title problem rather than a content problem, and cheap to fix.
- **Slipping** — pages that have drifted down over recent months.
- **Stale** — anything untouched for more than about eighteen months goes on the candidate list automatically, regardless of how it's performing.
- **Commercially important** — from {{commercial_priorities}}, pages that drive revenue deserve review far more often than the rest, on the order of quarterly.

**Per page: what specifically to change.** Not "update it." For each priority page, the concrete list — outdated statistics and what to replace them with, dead or redirected links, screenshots showing an old interface, advice that's no longer true given {{whats_changed}}, missing subtopics that readers now expect, a title that undersells, and structure that buries the answer.

**Substance, not signals.** Be explicit that changing the publish date, shuffling paragraphs, or adding a padding section does nothing and may cost trust. A refresh should add something a reader would notice: new data, a new section answering a question that's emerged since, a corrected recommendation, a real example. State what the reader gains for each page.

**The AI-visibility angle.** Recently updated, clearly structured pages are favoured when systems assemble answers, and rankings no longer reliably predict which pages get cited. So for each refresh, check that the core question is answered directly and early, that headings reflect how people actually ask, and that key passages make sense quoted alone.

**Merge or remove instead.** Pages where refreshing is the wrong answer — several of mine competing for the same query, or a page too thin to save. Say which to consolidate into which, and which to remove.

**Cadence.** Given {{capacity}}, a sustainable schedule: how many refreshes per month, how commercial pages and informational pages differ in review frequency, and how to keep this running rather than doing it once.

**Measurement.** What to record before each refresh so I can tell whether it worked, and how long to wait before judging.