Google has confirmed it rewrote search result links, organic and paid alike, into a server-side redirect format that hides the destination URL until after a click, and the rank-tracking industry says the change is already driving up the cost of reading a basic search results page.
The new links route through google.com/goto?url=[encoded destination] instead of pointing straight at the page you’re about to visit. Click one and Google 302-redirects you to the real destination, so the end result looks the same to a normal searcher. What’s gone is the hover preview, the little habit of checking a link’s actual URL in the status bar before you click it. That’s not a cosmetic tweak. It’s a basic transparency feature that millions of people, and a lot of SEO tooling, had quietly come to rely on. Here’s what happened, why it’s hitting rank trackers specifically, and what it means if your reporting dashboards start looking strange over the next few weeks.
What actually changed
The rewritten links were first spotted on June 23, 2026. Within two weeks, people documented them spreading across both organic results and paid ads, not just one or the other. By the first week of August, testing across different setups showed the goto format turning up widely enough that it clearly wasn’t a limited experiment. Then on August 26, 2026, rank-tracking vendor Nozzle reported something close to a full rollout, near 100 percent, across several residential IP providers it uses for testing. Google confirmed the change the same day.
Nozzle’s CEO, Derek Perkins, has been one of the more detailed voices tracking the rollout, and his numbers are the ones giving the SEO industry heartburn. Google has not published documentation for the goto parameter format. It hasn’t said whether this is permanent or a test. It hasn’t disclosed rate limits. And when asked directly, a Google spokesperson gave a statement that named no target and no specific abuse pattern:
“We have a long history of deploying technical measures against evolving forms of abuse, and we regularly take steps to protect our services and users,” a Google spokesperson said.
That’s it. No mention of scraping, no mention of SerpApi, no mention of AI companies building on search data. SEO commentator Barry Schwartz has floated his own read on the motive, suggesting the change looks like an attempt to make life harder for scraping tools generally, SerpApi specifically, and AI companies pulling search data at scale. That’s Schwartz’s inference, not anything Google has said on record, and it’s worth keeping that distinction clear given how much speculation this story is going to attract.
Why a redirect link breaks rank tracking
On the surface, a goto link redirecting to the same destination shouldn’t matter much. In practice, it changes the economics of reading a search results page for anyone who isn’t a human clicking one link at a time.
A rank tracker needs to know what URL actually sits behind each result. With a direct link, that’s free, it’s right there in the HTML. With a goto link, the real destination is encoded and can’t be decoded locally. The only way to find out where it goes is to actually follow the redirect. And Google blocks HEAD requests, the lightweight method that would normally let a tool check a redirect’s target without downloading a full response. That forces trackers into full GET requests for every single link they want to resolve.
Multiply that out across a results page and the math gets ugly fast. According to Perkins, resolving every link on a five-page set of search results, something a lot of rank-tracking workflows do routinely, can take 500 to 1,000 requests. Bandwidth was never the constraint here. Rate limiting is, and rate limiting is exactly the lever a company protecting its servers would want to pull.
Nozzle’s response during the test window was blunt: it discarded goto-only results pages rather than try to resolve all of them. Other tools reported similar strain. Places Scout reportedly saw its ranking reports affected. A tool called Serpdino reportedly started hitting verification interstitials on standard search URLs, pages that return a 200 status and the correct title but no actual results container, which is its own kind of silent failure. SERP Lens users reportedly made the same call Nozzle did and discarded affected data rather than report numbers they couldn’t trust.
This is the second cost shock in under a year
Rank tracking vendors were already dealing with a version of this problem. In September 2025, Google removed the num=100 parameter that let tools pull 100 results in a single request. Overnight, anyone tracking deep rankings had to make ten requests where one used to work. Semrush confirmed a tenfold jump in operational cost within a day of the change. AccuRanker responded by limiting its own tracking to the top 20 results. SerpApi’s workaround capped out at three organic results per request. Ahrefs eventually restored top-100 tracking by the end of October 2025, but quietly, without ever disclosing how it managed to do it.
The goto rollout compounds that earlier change rather than replacing it. Removing num=100 multiplied how many pages a tool has to fetch to see a full results set. The goto format multiplies how many requests it takes to make sense of a single page once you’ve fetched it. Stack the two together and the cost of comprehensive rank tracking has gone up by a lot more than either change alone would suggest.
The SerpApi lawsuit is the backdrop here
None of this is happening in isolation. Google sued SerpApi in December 2025 under DMCA Section 1201, accusing the company of circumventing its SearchGuard anti-bot system. A judge dismissed most of the claims in July 2026, which looked like a real setback for Google’s case. Google didn’t let it go. It filed an amended complaint in August 2026, dropping some of its original theories but keeping the fight alive, with a hearing now set for September 29, 2026.
That timeline overlaps almost exactly with the goto rollout. Whether the two are connected is, again, Schwartz’s inference and not a confirmed fact. But it’s hard to look at a company simultaneously suing a scraping tool in court and making scraping technically harder in production and conclude the two efforts are unrelated in spirit, even if Google has never said so directly.
What this means if you run SEO reports for a living
If you manage rankings for a site, an agency, or a client portfolio, there are two practical things worth doing right now. First, expect your rank-tracking vendor to adapt the same way AccuRanker and SerpApi did after num=100: price increases, reduced tracking depth, or gaps in coverage that show up as missing data points rather than a clean error message. Ask your vendor directly how they’re handling goto links specifically, and whether that changes what you’re paying or what depth you’re getting.
Second, be skeptical of sudden ranking volatility in your dashboards over the next month or two. If a dashboard shows a page cratering from position 4 to nowhere, that might be real, or it might just be a tracker that lost visibility into a results page and is reporting a gap as a drop. Data gathered before, during, and after this rollout isn’t directly comparable, the same caveat that applied after num=100 broke historical trend lines for a while.
This also fits a pattern that’s been building for a year. Reddit CEO Steve Huffman already blamed Google’s AI Overviews for falling search traffic to publisher sites, and Penske Media has taken its fight over AI Overviews to the court of appeals in an antitrust case against Google. Different mechanisms, same underlying story: publishers and the tools that serve the SEO industry are losing visibility into, and leverage over, how Google Search actually behaves, one layer at a time.
The bottom line
Google is well within its rights to defend its infrastructure against scraping, and nobody should expect SearchGuard-style defenses to stop evolving. But rolling this out without documentation, without stated rate limits, and without naming what it’s actually defending against puts every rank-tracking vendor in the position of guessing at the rules while their costs climb. If you’re an SEO relying on third-party rank data right now, treat anything unusual in your reports with suspicion until your vendor tells you exactly how they’re handling goto links. The real risk isn’t that tracking gets harder, it’s that nobody tells you when the numbers you’re looking at quietly stopped meaning what they used to.
The Teck Hustlers Editorial Team covers technology, business, finance and gaming news for readers who want the story behind the headline. Every article is researched, written and edited by our in-house team, checked against company announcements, filings and reputable reporting, and updated when new facts emerge. We focus on what each development means for founders, professionals, investors and everyday users. Have a tip or a correction? Reach us through our contact page.
