Google Search

Google Publishes Help Doc on Allowing Google-Agent Traffic

Google has published a new help document explaining how to allow Google-Agent, its agent bot, to access your website. Search Engine Roundtable reported the document on November 17, 2026. The document defines Google-Agent as the name used by agents hosted on Google infrastructure to navigate the web and perform actions upon user request. In plain terms, this is the bot that does things on behalf of users: booking, buying, filling in forms, and completing multi-step tasks. Google is now telling site owners how to let it in, which means the agent web is moving from experiment to infrastructure.

What Google-Agent is

Google-Agent is not Googlebot. Googlebot crawls pages to index them for search. Google-Agent visits pages to act on them, carrying out tasks a user asked for. The distinction matters because the two bots need different permissions. A site that blocks all bots except Googlebot could be blocking the very agent that would have completed a purchase or a booking for one of its visitors. The new help document exists to prevent that kind of accidental block, and to give site owners a deliberate choice about agent traffic.

The timing is not a coincidence. Google made its Gemini Tasks agent free for hundreds of millions of users on November 17, the same day the help document appeared, and that agent can call local businesses, research products, compare options, find discount codes, and complete purchases using Google Pay. Those tasks happen on the real web, on real sites, through an agent acting for the user. If sites block the agent, the tasks fail. If sites allow it, the agent becomes a new kind of visitor, one that arrives with intent already formed and ready to convert.

Why the document matters now

The agent web is arriving fast, and the rules for it are being written in real time. Until now, site owners have had to guess how Google's agents identify themselves, what they can do, and how to control them. A help document from Google is the first official answer to those questions, and it signals that Google expects agent traffic to become a normal part of web traffic. Sites that plan for it now will be ready when agents start arriving in volume. Sites that ignore it may find their pages blocked to agents by default, or their forms and checkouts broken for the fastest-growing segment of visitors.

There is also a defensive angle. The same week Google published this document, it was testing sign-in walls to block scrapers from its search results, and it was still fighting in court over SerpApi's scraping of Google results. Google is drawing a line between agents it hosts and controls, which get a help document and a welcome, and scrapers it does not control, which get sign-in walls and lawsuits. Site owners should expect that line to get sharper, not softer.

What it means for SEOs and link builders

For SEOs, the immediate task is to check how your site treats agent traffic today. Review your robots.txt, your firewall rules, and any bot management settings, and confirm that Google-Agent is not being blocked by a rule written for a different era. If your site has a checkout, a booking flow, or a form that an agent might complete, test it with agent traffic in mind. A page that works for a human visitor but breaks for an agent is a page that will lose conversions as agent usage grows.

For link builders, the agent web changes what a good link is worth. A link on a page that an agent visits and acts on is a link in a conversion path, not just a ranking signal. Content that helps an agent complete a task, such as clear pricing, accurate availability, and straightforward product information, becomes more valuable, because the agent reads it and acts on it directly. The pages that win in the agent era will be the ones that are easy for both humans and machines to understand and use, and the help document is Google's first hint about how to build for that future.

What to watch next

Watch for Google to publish more detail on how Google-Agent identifies itself, how site owners can verify it, and whether it respects the same robots.txt rules as Googlebot. Watch too for the first reports of agent-driven conversions showing up in analytics, and for whether agents start appearing as a distinct traffic source. The help document is a small step, but it is the first official one, and the sites that read it now will be ahead of the ones that wait for the traffic to arrive.

Sources

Sources: Search Engine Roundtable, November 17, 2026.

Related reading on this site: November 2026 Core Update Still Rolling Out as Volatility Peaks on November 14, Google Tests Temporary Chats in AI Mode, Borrowing the Feature From Gemini, Mueller: Crawled or Discovered, Not Indexed Is a Quality Signal, AI Overviews Showed Up in Only 20% of Local Searches, and Mostly Copied the Map Pack, Semrush Rebuilds Traffic and Market Data With a New Model, and Old Numbers Will Change, and Domain Authority Barely Correlates With AI Citations, Search Atlas Finds.

More from Linkbaits

Turn this playbook into links

Linkbaits helps you create content that earns citations and backlinks. Start with the toolkit and build your first linkable asset this week.

Start free trial