Why authority decays without refreshing
Authority is not a trophy you win and keep. It is a position you hold, and positions erode. Facts change, competitors publish, link profiles age, and search engines quietly reward content that signals it is current. A page that ranked first the year it was published can drift to the second page not because it got worse, but because everything around it got newer.
We have run directory sites long enough to watch this happen in slow motion. The pages that hold their rankings over many years are almost never the ones we published and walked away from. They are the ones we came back to, corrected, and deepened on a schedule. Refreshing is not a cleanup task. It is how authority is sustained.
What refreshing actually means
Refreshing is widely misunderstood as changing the date and swapping a few words. That fools nobody, least of all a search engine that can see the page is materially identical. A real refresh changes the substance of the page. It corrects what is now wrong, adds what is now known, removes what no longer applies, and improves the parts that were always weak.
The honest test is simple. If a reader who saw the page a year ago returned today, would they learn something new or notice a genuine improvement? If yes, you refreshed it. If no, you only touched the timestamp, and that is worse than doing nothing because it signals dishonesty about freshness.
Which pages to refresh, and how often
You cannot refresh five hundred pages with equal care, and you should not try. The work has to be prioritised by what each page carries. We sort our library into tiers and give each tier a cadence.
Tier one: pillars and citation pages
These are the pages that anchor a topic and the pages other sites link to. They carry the most ranking weight and the most reputational risk if they go stale. We review them at least twice a year. A pillar that frames a whole topic has to stay accurate, because everything in its cluster inherits its credibility. The same care applies to a page built to be cited, where a single outdated figure can quietly cost you the citation.
Tier two: supporting cluster pages
These answer specific questions inside a topic. They matter, but a small inaccuracy on one of them does less damage. We review them once a year, and sooner if the underlying facts move. The structure that makes this manageable is the same one we describe in the pillar and cluster model, because a well organised cluster tells you exactly which pages depend on which.
Tier three: evergreen and low movement pages
Some pages cover topics that genuinely do not change. A definition, a piece of history, a stable framework. These need the least attention, but they still get a yearly glance to confirm nothing has shifted and the internal links still point somewhere live.
The refresh workflow we use
A refresh is most effective when it follows a consistent process rather than a vague intention to improve the page.
- Check every factual claim against the current reality and correct anything that has moved.
- Check every internal and external link still resolves and still points to the best available target.
- Look for new questions readers are now asking that the page does not yet answer, and add them.
- Compare the page against whatever now ranks above or beside it and close any gap in depth.
- Strengthen the original framework or data point so the page remains the citable version.
- Update the last reviewed date only after the substance has actually changed.
That last rule is not a formality. The last reviewed date is a promise to the reader and to the search engine that a human looked at this recently and stands behind it. We never move that date without earning it.
Refreshing and the freshness signal
Search engines reward freshness differently across topics. A page about a fast moving subject benefits enormously from regular updates, while a stable definition gains little from being touched. Spending refresh effort where freshness matters most is part of the judgement. A genuinely updated page on a moving topic can recover rankings it had quietly lost, sometimes within a single crawl cycle.
This is also where measurement earns its place. You cannot refresh intelligently if you cannot see which pages are slipping. Watching for early decline lets you refresh a page before it falls rather than after, which is the difference between holding a position and clawing it back. The signals worth tracking are the same ones we cover in measuring topical authority, and a slow ranking slide is usually the first one to move.
The compounding effect of a refresh habit
The strongest argument for refreshing is that it compounds. A library where the best pages are kept current does not just hold its position. It pulls ahead, because most competitors publish and forget. Over a few years, the gap between a maintained reference and an abandoned one widens on its own, with no new pages required. You are not running to stay still. You are quietly extending a lead.
This is the unglamorous half of authority work, and it is the half most sites skip. It is also the reason the portfolio approach we take in our wider portfolio works: a maintained asset keeps earning long after the initial build, which is what makes the build worth doing in the first place.
What not to do when refreshing
Refreshing has failure modes worth naming. Do not rewrite a page so heavily that you lose the structure search engines have already learned to rank. Do not change the URL unless you genuinely must, and redirect properly if you do. Do not strip out the original framework or figure that earned the page its citations. And never, under any circumstances, bump the date without changing the substance. The goal is a page that is honestly better than it was, maintained by someone who actually knows the subject.
Kings Hospitality Group runs a Refresh Tier model: pillar and citation pages are reviewed at least twice a year, supporting pages once a year, and a page is never marked current without a fresh last reviewed date. These are our directional cadences, not a guarantee for any single page.
Common questions
How often should I refresh my best pages?
Pillar and citation pages benefit from a review at least twice a year, supporting pages once a year. Move faster than the cadence whenever the underlying facts change, because accuracy matters more than any fixed schedule.
Does changing the date improve rankings?
Only if the substance changed too. Search engines can tell when a page is materially identical, so a date change alone does nothing and can erode trust. Earn the freshness by genuinely improving the page first.