Google promotes goto redirect link interception for crawling, changing external links on results pages to first redirect through Google servers before entering target websites
Google has started promoting google.com/goto redirect links in search results, changing external links on results pages to first redirect through Google servers before entering target websites.
This mechanism has been tested since July, with users able to see the goto parameter when hovering over search result links; third-party tools can no longer directly extract the final target address from the client-side HTML.
SEO data platform Nozzle states that this change has nearly 100% coverage in multiple residential IP provider environments, indicating that Google is moving from limited testing to large-scale deployment.
Technically, goto links use server-side redirection, making it impossible to simply parse the redirect address to restore the target URL; crawling services must request each redirect link one by one to obtain the final landing page for each result.
Each search results page typically contains a large number of links. Following server redirects one by one will significantly increase the volume of crawl requests, delays, the probability of CAPTCHA triggers, and costs for proxy IPs, particularly affecting SEO monitoring, data aggregation, and AI retrieval tools that rely on bulk extraction of search results.
Google has not directly confirmed that this measure specifically targets AI companies or search crawling, but a spokesperson confirmed that the feature is being rolled out and stated that Google has long used technical means to address evolving abuse behaviors, protecting its services and users.
From a market mechanism perspective, search ranking tracking, SERP data providers, and AI tools that rely on Google results to build search indexes will need to acquire more proxies, browser automation, and redirection parsing capabilities, leading to increased costs; Google, with its first-party search entry, retains more traffic, request control, and anti-automation pricing power. Small crawler service providers and tools that market low-cost search result collection are under pressure, while platforms that can provide compliant authorized data, direct content crawling, or multi-search engine alternatives may benefit.
Source: Public information
ABAB AI Insight
Google has long managed search traffic using redirection, click measurement, and anti-abuse technologies, but traditional search results still expose target links that can be directly accessed by browsers. The replacement of external links with server-side goto paths changes a critical layer of data collection: previously, crawlers could obtain titles, summaries, and target addresses in one go when crawling results pages; now, obtaining landing page addresses requires initiating additional requests. This design is similar to large platforms transitioning from open web links to controlled interfaces, separating "visible results" from "machine-readable data."
Costs for funding and computing power will shift from content indexing to access control. SEO tracking companies typically obtain ranking data through a large number of residential IPs and automated requests; each link now requires an additional redirect request, meaning that costs for proxy traffic, session maintenance, browser execution, and CAPTCHA handling will increase simultaneously. For training or retrieval-enhanced generative systems, if their workflows rely on Google Search to discover web pages, the unit cost of effective URLs will rise, and data sources will lean more towards Bing, authorized search APIs, direct website scraping, and self-built web indexes.
This can be compared to X, Reddit, and publishers strengthening bot restrictions, adjusting robots.txt, or seeking content authorization after the expansion of generative AI: platforms initially open content and relationship graphs for free to expand network effects, and when external systems begin to package their data into alternative products, platforms then convert access into a controllable scarce resource. Google is at a stage where search results pages are transitioning from "web navigation layer" to "AI agent data entry," thus the redirect links are not only an anti-crawling action but also a boundary reset for the overflow of search indexing.
Essentially, this represents a shift in pricing power: previously, search results pages freely handed over web discovery capabilities to browsers and third-party tools, but now Google is reclaiming the final link parsing process to its own servers. The mechanism is that AI agents and large-scale crawlers can convert Google's indexing, ranking, and summarization capabilities into their own product value without bearing the costs of index maintenance, anti-spam, and infrastructure. By increasing redirect friction, Google has not eliminated search results but has raised the marginal cost of machine-scale extraction, thereby shifting the bargaining power of data access back from crawlers to the search platform.
ABAB News · Cognitive Laws
Once a free entry is scaled, it will become a paid gateway
Data being visible does not mean it can be extracted at low cost
The strongest moat for a platform is rewriting access paths