Breaking
Vodafone Zoozoos: The Silent Ad Mascots That Took Over IndiaThe Global Best Ads This Week: 4 Ideas Worth StealingTaika Waititi’s steamy ad makes breast checks a shower habit82% forget breast checks. Taika Waititi's fix: the showerGoogle's YMYL: the hidden trust filter behind 1B health searchesVodafone Zoozoos: The Silent Ad Mascots That Took Over IndiaThe Global Best Ads This Week: 4 Ideas Worth StealingTaika Waititi’s steamy ad makes breast checks a shower habit82% forget breast checks. Taika Waititi's fix: the showerGoogle's YMYL: the hidden trust filter behind 1B health searches
SEO

Google says new pages take about 20 hours to be found

Google's internal timing data shows a new URL is typically discovered in about 20 hours—but slow cases stretch to weeks, months, or never. Here's what to fix.

Google's 20-hour page-discovery window

Marketers have spent years guessing how long Google needs to find a new page. On October 2, 2026, at Search Central Live Deep Dive Europe in Barcelona, Google analyst Gary Illyes put internal timing data on the screen. The headline: a new URL is typically discovered in about 20 hours. The slowest cases run from weeks to never.

The numbers come from Google’s internal analysis, as recapped by John Campbell of the agency ROAST. One nuance from the recap: the live session discussed fastest, typical and slowest times, but the published tables mostly show the typical and slowest columns.

Google’s timing ranges

The figures split into three buckets: crawling, indexing and serving.

  • New URL discovery: about 20 hours typical, with slow cases of weeks to never.
  • Indexing end to end: about 1.5 hours typical, but site moves land at 1–3 months and can stretch past a year.
  • Core update recovery: 3–6 months typical, or up to a year if a site has to wait for the next core update.

Robots.txt changes are far quicker at roughly a day, while manual action removals fall around 1–2 weeks. The real insight is not any single number but how delays stack.

Why ‘never’ is a quality signal

Illyes noted that many processes are linked: a page cannot be indexed until it is crawled, so each slow stage compounds. The recap’s commentary points out that ‘never’ shows up often and is usually tied to quality. In practical terms, technical speed does not rescue a page Google judges not worth the crawl.

This reframes the old crawl budget conversation. Google runs one central crawling system shared across Search, Ads, Shopping and Images, balancing host load with demand. If quality is low, no amount of sitemap hygiene changes the outcome.

What to do with these numbers

The value for marketers is a set of Google-supplied ranges to replace gut feel and agency folklore. Compare them with your own Search Console reports before a migration, a core update analysis or a new content push.

Start with the fastest layer: check robots.txt, rendering and server errors that can block Googlebot. Then separate discovery speed from indexing speed. A page found in 20 hours is not the same as a page earning a place in serving.

Finally, set expectations around the long horizons. Site moves are months, not days. Core update recovery is measured in quarters. When Google says a process can end in ‘never’, the fix may be content value, not code.

That practical lens matches Illyes’ closing line from the three-day event: “AI on Google is just SEO.” The same crawl, index and serving logic now underpins AI search.

Source: PPC Land

Leave a Reply