Google's /goto redirects, and what they actually change

For years a Google Search result did something very simple. You clicked a link, and Google sent you to the website.

Now Google is inserting itself into the journey. Instead of linking directly to a page, some Search results route the click through a Google URL resembling:

google.com/goto?url=...

The user still reaches the intended page. To most people the change is close to invisible.

The systems that observe Search are another matter. Rank trackers, SEO platforms, data collectors and AI tools all read that link, and for them the difference is significant.

Google has not changed where the result leads. It has changed how easily the destination can be observed. That is a smaller thing than it sounds, and it tells you more than it looks like it should.

What has actually changed

Historically the link attached to a result exposed the destination URL directly. A tool reading the results page could identify the ranking page from the link itself.

With the new approach, the visible link can point first to a Google /goto address. Google then resolves that address and redirects the browser onward.

Google has confirmed the rollout is intentional, describing it as part of the technical measures it deploys against evolving forms of abuse and to protect its services and users. It has not publicly said which specific abuse this targets.

The rollout appears to have expanded considerably after testing earlier in 2026. Nozzle's Derek Perkins reported seeing /goto-only results on almost all searches across several residential IP providers, though that is one observation and should not be read as a universal rollout figure for every user, country and result type.

SerpApi has reported the format appearing beyond conventional organic listings, including Search, News, Videos and AI Overview-related responses. Its engineers say the destination remains encoded inside the redirect data, but that resolving the different implementations takes additional work.

In plain English: humans still follow the link easily. Machines can no longer read every destination quite so cheaply.

Is this a ranking update

No, on the evidence available. There is nothing to suggest /goto redirects change rankings, crawling, indexing or the relevance signals that order results.

They alter the delivery of an outbound click from the results page. They do not appear to alter how Google evaluates the page itself.

That distinction matters, because SEO has a habit of turning every Google change into a ranking emergency. This one is a measurement and infrastructure issue. Your content has not become worse. Your rank tracker may simply have become less certain about the URL it is looking at.

So what does it affect

Some third-party SEO data may become less reliable

A rank tracker may previously have collected destination URLs straight from the results page. If it now receives an opaque Google redirect, it has to follow or decode that redirect before it can identify the final page.

Doing that once is straightforward. Doing it reliably across hundreds of links, result types, locations, devices and changing formats is an engineering problem.

SerpApi says the change has affected rank tracking, source extraction, link analysis, content discovery, AI and retrieval pipelines, and anything that stores or compares destination URLs. It is adapting its resolution process as the implementation evolves.

This does not mean every SEO tool is now wrong. It means you should be alert to discontinuities: missing URLs, sudden landing-page changes, unexplained ranking gaps, unusual disagreement between two tools that used to agree.

The cost of observing Google Search has gone up

Google says the change is an anti-abuse measure. Frustrating large-scale scraping is likely to be one of its practical effects, which is an inference rather than a statement: a collector that once read a destination from the page may now need to resolve a separate redirect for every result.

It will not stop sophisticated providers. It does add computational cost, latency, failure points and maintenance.

That is often how platforms restrict machine access. Not by closing the door, but by making it more expensive to open at scale.

AI systems are part of the story

Modern AI search tools and agents need more than titles and snippets. They frequently need to identify, retrieve, compare and cite source pages.

If the result URL becomes harder to resolve, anything built on scraped Search results gains another dependency and another point of failure. SerpApi lists citation and source extraction, AI agents and retrieval-augmented generation pipelines among the affected uses.

This does not mean Google has blocked AI systems from reaching your website, and it does not prove your visibility in AI Overviews will change. It means third-party systems using Google's results as a discovery layer have to work harder to reach and verify the source. That is a subtle difference, and an important one.

Google is becoming both the marketplace and the measurement authority

Here is the larger point.

Google decides what appears in its results. Google now controls another part of the route from result to destination. And when third-party observation becomes less stable, marketers get pushed further towards Google's own reporting.

There is no indication that Search Console itself is disrupted. It measures activity inside Google's systems rather than depending on an external scraper reading result links. But Google has not specifically guaranteed that the rollout has no effect on every Search Console report, so treat "unaffected" as a strong expectation rather than a newly confirmed promise.

Will this change attribution in GA4

There is no broad evidence at present that ordinary Google organic visits are being reclassified because of /goto.

Browsers pass referral information as a user moves between pages, and redirect behaviour can affect attribution depending on the implementation. Google Analytics documents how intermediary referrals influence attribution and how referral exclusions work.

But possibility is not proof of an observed problem. So the sensible response is not to declare an attribution crisis. It is to measure:

• Whether Google organic sessions change sharply without a corresponding change in Search Console clicks.
• Whether unexpected google.com referral classifications appear, or direct traffic rises.
• How landing-page sessions, Search Console clicks and conversions compare before and after the rollout.
• Whether anomalies survive investigation before anyone changes a channel report or a budget.

Measure first. Panic later, if required.

We are running exactly that comparison across the accounts we operate, and we will publish what it shows, including if it shows nothing.

What to do now

Keep doing SEO

There is no reason to rewrite your content strategy, change canonical tags or optimise for /goto. Useful content, sound technical implementation, clear site structure, pages worth visiting. The fundamentals are unchanged.

Validate third-party tools against first-party evidence

If a rank tracker reports a sudden change, compare it against Search Console impressions, clicks, average position and the pages actually receiving organic traffic. One dataset should not be allowed to declare an emergency on its own.

Ask your vendors a better question

Do not ask only whether they have fixed the Google redirect issue. Ask:

"Resolved" is a product claim. Data completeness is a measurement.

Build decisions from more than one signal

Rankings tell you whether a page appears. Search Console tells you whether it earns impressions and clicks. Analytics tells you what visitors do. CRM and commerce data tell you whether any of it produced a customer.

None of those is the whole truth. The closer a decision sits to revenue, the less sensible it is to base it on a single platform or a single tool.

Access is not ownership

Marketers talk about "our Google data" as though it were an asset sitting safely inside the business.

Much of it is not. It is access, granted through an interface, a report, a result page or a third-party tool whose own access can change overnight.

The /goto rollout is a small technical change carrying a larger message. The systems through which we observe marketing are not neutral pipes. They are controlled environments, and the owner of the environment sets the rules, the visibility and the cost of access.

So the question is not really whether Google's new redirect will affect your rankings. Probably not, on what we know today.

The better question is how many of your marketing decisions depend on data you cannot independently verify.

That is the part worth fixing.

Sources

• Google confirms deploying goto URL redirects to search results links, Search Engine Land
• Confirmed: Google Search rolling out google.com/goto tracking parameters, Search Engine Roundtable
• Google's new /goto redirect URLs: resolution in progress, SerpApi
• Google's new "Goto" URL redirects: what it means for SEO, Rankmetry
• Referer header, MDN Web Docs
• Identify unwanted referrals, Google Analytics Help