Why does it work so well?
A new article starts from zero: no history, no links, no position. Google keeps it under observation for weeks before ranking it seriously. An existing page that has already ranked owns all that capital: age, internal and sometimes external links, a click history. When it slips, it is almost never because it turned bad, it is because competitors published something fresher and more complete.
Bringing that page back up to standard costs two hours. Getting the same result with a new page costs months. That is the whole argument, and it is massive: updating is the only content lever with a near-immediate effect, since Google recrawls a modified page within days.
Which pages should you refresh?
In Search Console, Performance report, compare the last 3 months to the previous 3 months and look for three profiles:
- The sliders. Stable impressions, declining clicks, a position drifting from 4-6 to 8-15: the page is losing ground to fresher content. Top priority.
- The almost-theres. Average position 5 to 12 for months, never in the top three: they are missing one small thing, often completeness or internal linking.
- The dated ones. Any content whose title or text mentions a past year, an outdated price, a tool that no longer exists. They erode the trust of both the reader and the engine.
Rank them by commercial potential: a slipping service page comes before a news post from 2023.
What should you change, exactly?
- Update the facts. Numbers, prices, dates, screenshots, tool names. That is the visible face of freshness, for the reader and for the engine.
- Move the answer up. If the page makes people wait 400 words before answering the question, rewrite the opening: answer first, details after. That is also what makes the page citable by AI.
- Fill the gaps. Type the target query, open today's top three results, note the subtopics they cover and you do not. Add the missing sections, with your examples and your numbers, not theirs.
- Strengthen the internal links. Add links from your strong, recent pages to the refreshed page, and from it to your service pages.
- Polish the title and meta description. A title rewrite alone can lift the click-through rate by several points.
What should you not touch?
- The URL. Never. It carries all the history. If you absolutely must change it, a 301 redirect is mandatory, but avoid it.
- The target query. Refreshing is not refocusing. If the page targets a different query after your edits, you are starting from zero again, with a bonus risk of keyword cannibalization with another page.
- The publication date, without a real update. Changing the date without changing the content gets detected and ignored. Instead, show an honest "Updated on" note backed by genuine changes.
- Whatever already ranks in positions 1-3. You do not fix what is winning. A light touch-up of dated facts, nothing structural.
Measure the effect
Note the page's position and clicks on the day of the update, then check again at 2, 4 and 8 weeks in Search Console. The recrawl happens within days, the position effect in 2 to 6 weeks. If nothing moves after 8 weeks, freshness was not the problem: look at the query's competition and the internal linking. A healthy rhythm for a small business: one update per week, alternating with new content. That is the regimen that keeps the whole website alive without burning anyone out.