Barry Schwartz at Search Engine Roundtable published two documented cases this week of Cloudflare settings cutting off Googlebot entirely, each one costing a site weeks of rankings and traffic. The failure leaves no trace in Google Search Console, so a site can pass every on-page and robots.txt check and still disappear from the index. Both cases came from working SEOs posting on LinkedIn: Jonathan Bird described one incident and Brodie Clark described a separate, unrelated one, and in both a specific setting the team had turned on, not a Cloudflare default, was responsible.
In Bird’s case, a managed IT provider flipped on a Cloudflare setting meant to stop every crawler from reaching the site, Googlebot included. Bird wrote that the site’s traffic collapsed for two weeks before the cause was found. Google Ads kept charging the account even as the crawl block cut off traffic, the site’s Merchant Center listings vanished, and organic search traffic disappeared. Bird said the business was only beginning to recover as he posted the account on LinkedIn.
Clark’s case involved a different site: an online marketplace that is one of several domains the same company operates. He said the visibility collapse looked like it could be blamed on a core update or a spam action, but the real cause was operational. Bots were overloading the marketplace’s servers, so the team added firewall rules restricting bot access to keep the site running. Those rules blocked Google’s crawler as a side effect, producing damage severe enough to be mistaken for an algorithm penalty.
Schwartz noted the underlying problem predates Cloudflare by decades: site owners have blocked Google via Apache rules or robots.txt errors for more than twenty years. What has changed, he added, is the sheer number of overlapping tools and settings now sitting in front of a typical site, which makes an accidental block easier to introduce and harder to trace.
This publication flagged the same underlying risk on August 5, when a single reported case pointed to Cloudflare’s AI bot toggle as a possible cause of lost Googlebot access. Bird’s and Clark’s cases widen that pattern beyond one toggle. Bird’s block came from a general crawl-control setting, and Clark’s came from firewall-level bot rules, showing the failure reaches multiple Cloudflare controls rather than a single feature.
For a search team, the operational risk is that this failure mode produces no signal inside Google’s own tools. A site can be fully compliant at the CMS and robots.txt level and still vanish from the index because an edge rule several layers upstream, at the CDN or firewall, is intercepting Googlebot before it reaches the server. Teams running Cloudflare or any comparable bot-management layer should treat a live Googlebot fetch test as a standing check on every new firewall or crawl-control rule, not a one-time setup step.
Barry Schwartz reported these cases for Search Engine Roundtable on August 13, 2026, citing LinkedIn posts from Jonathan Bird and Brodie Clark.
