Publishing a new post is starting from zero: no links, no history, no queries where Google has already tested you. Refreshing an existing post starts from whatever that post already earned. That's the entire argument, and the data keeps agreeing with it.
Why old pages refresh well
A page that ranks 11-15 has already passed the hard tests. Google has indexed it, matched it to dozens of queries, and found it credible enough to show on page two. What it lacks is usually not authority but freshness and completeness. The information is stale, a competitor answered a sub-question better, or the format no longer matches what searchers click.
A refresh fixes all three without touching the URL's accumulated equity. The links it earned still point to it. The behavioral history survives. You're renovating a house in a good neighborhood instead of building on empty land across town.
What a real refresh includes
Deleting anything outdated — not just adding. New sections for questions the original ignored (People Also Ask is a free syllabus). Updated statistics with visible dates. Tightened intro so the answer lands before the scroll. New internal links in both directions. If images are ancient screenshots, re-shoot them. Then, and only then, update the modified date — readers can smell a fake date, and so can evaluators.
How to pick refresh targets
Rank for the refresh, not the deletion: pages ranking 5-30 on queries with real impressions. Position 5 to 8 is your fastest money — a refresh often nudges these into the top three. Position 25 and sinking? Better as a merge into a stronger URL. And pages with zero impressions for a year aren't refresh candidates; they're deletion candidates.
Publishing cadence feels productive. Improving cadence is productive. A site that refreshes two strong pages a week will typically outrun a site that publishes two thin new ones — with half the effort and none of the new-page gamble.
