Google Cloaks Search Result URLs Behind /goto Redirects to Choke Scrapers

Google Search has begun swapping the real destination URLs in organic results for opaque google.com/goto?url=... links. The url parameter isn’t plain base64 or a readable query string like the older google.com/url wrapper — it’s a Google-specific encoding that appears to reference an internal index record and can’t be decoded offline. To learn where a result actually points, a client has to hit the /goto endpoint and read the destination from the HTTP Location header rather than following the redirect through to the page. As of late August 2026, this behavior is consistent across logged-out and private-browsing sessions, suggesting it has moved past a narrow experiment.

The motivation is economic: with plaintext links, a scraper could harvest thousands of destination URLs from a single HTML page without ever contacting Google again. Under the /goto scheme, every result requires a round trip back to Google just to resolve its target, making bulk extraction slower, noisier, and easy to fingerprint when one client resolves hundreds of links in sequence. The change lands alongside earlier friction — dropping the &num=100 parameter and tightening BotGuard/SearchGuard — as part of a sustained effort to raise the cost of automated SERP harvesting, particularly by AI crawlers and SEO tools building competing indexes.

For teams that depend on search data, the practical takeaway is that naive HTML parsing of Google results is increasingly broken. The article’s author, SERP-API vendor Autom.dev, says it has updated its pipeline to resolve /goto links server-side and return final destination URLs in the same structured fields customers already consume — a reminder that this arms race mostly shifts the resolution burden onto intermediaries rather than eliminating scraping outright.