
CDN migration: why Googlebot hits your origin server harder
Abondance reports that faster CDN response times can lead Google's robots to raise a site's crawl capacity limit, increasing crawl volume soon after a migration. Because a CDN starts with an empty cache, each first request on a URL goes back to the origin server.
Abondance reported on 16 November 2026 that a migration to a CDN can produce a counter-intuitive side effect: Googlebot may start requesting your origin server more heavily, not less. According to the report, Google states that faster response times from a CDN encourage its robots to progressively raise the crawl capacity limit set for the site, which can increase crawl volume shortly after the switch. The second point in the report is mechanical: a CDN starts with an empty cache, so every first request on a URL travels back to the origin.
The mechanism, and what it does not say
Two effects combine. The CDN makes responses faster, and Google's crawler is said to adapt its crawl capacity accordingly. At the same time, the cache has nothing in it, so the first hit on each URL is a cache miss served by the origin. Neither effect is a penalty or a malfunction; both follow from the setup itself.
The report does not specify the size of the crawl increase, the delay over which it appears, or which origin configurations are most exposed. Any figure circulating on this topic should be treated as unverified. What the report does establish is the direction of the effect and the reason for it: speed gains translate into more crawling, and cold cache translates into more origin traffic.
What an SEO or platform team can do this week
If a CDN migration is planned or recent, the origin server deserves a capacity review before the switch rather than after it. Concretely:
- Check whether the origin can absorb a burst of cache-miss requests, especially on large catalogues, paginated archives and faceted URLs.
- Look at how much of the site is actually cacheable, and whether headers let the CDN store HTML pages rather than only static assets.
- Review crawl statistics and origin logs separately before and after the migration, since the origin now only sees misses.
- Decide deliberately how much of the site should be open to crawling, rather than letting the capacity limit do that work alone.
Why it matters beyond the server
For teams whose traffic depends on organic search, a CDN is usually framed as a performance project. The report suggests it is also a crawl-budget project. A faster site is crawled more; a cold cache sends that crawling to the machine you were trying to protect. Treating origin capacity and crawlable surface as one decision, taken before migration, is cheaper than fixing an overloaded server once the crawler has already scaled up.
There is no universal recipe in what Google told Abondance, and no published threshold to aim for. The practical takeaway is to instrument before and after, and to know exactly which URLs a cold cache will push to the origin in the first days.
Also worth reading
-
Google AI Mode upgrade: what complex queries mean for search traffic
Google is upgrading AI Mode with Gemini 4 Hydrogen, promising better handling of complex queries, more relevant links and richer visual answers. For teams that depend on organic search, the practical question is whether those links and visuals still send visitable traffic.
-
SEO, GEO and Google Ads: what SMEs must reassess for 2027
FrenchWeb reports that SMEs can no longer treat search visibility as a matter of rankings alone, as AI-generated answers and automated advertising reshape discovery. The central question becomes how many actual customers each channel brings in.
-
Google ranking: RESONEO maps 305 signal groups in its code
RESONEO says it has catalogued 305 groups of ranking signals, recorded in a tool called RankLab, split between signals that apply to each result and each query. The report, relayed by Abondance, does not document how the signals are weighted or measured.



