Google’s /goto Redirect: What It Means for Rank Tracking, Reporting, and Your Agency

First published 1 September 2026. This story is developing, and tool responses and GA4 attribution data are still emerging. We will update as new information comes in.

Hover over a link in Google search results today. You won’t see a URL pointing to a website. You’ll see something like google.com/goto?url=…, an encoded redirect that routes through Google’s own servers before sending you anywhere.

Most users won’t notice. But the change behind that link may affect your clients’ reporting, namely keyword ranking data, SERP visibility reports, and even organic traffic attribution in GA4.

What the /goto redirect actually is

In late July 2026, SEOs started noticing that Google had stopped including direct destination URLs in search result links. Instead, every result now routes through google.com/goto, a server-side redirect that encodes the target URL using Protocol Buffers (protobuf). By late August, Derek Perkins of Nozzle confirmed it had reached near-100% rollout across residential IP providers.

The destination URL cannot be decoded locally from the redirect. To retrieve it, a tool has to follow the full redirect chain through Google’s servers, and that’s where the operational problem begins.

Google confirmed the change with a statement that told us everything and nothing: 

Before this change, rank trackers read destination URLs directly from SERP HTML. That’s gone. Now, every keyword tracked, every SERP scraped, every position checked requires the tool to follow a redirect chain to retrieve the URL it used to get for free. According to Nozzle, resolving all links across a single five-page ranking check now requires 500-1,000 HTTP requests. Google doesn’t allow HEAD requests through, so there’s no lightweight workaround. The cost compounds per keyword, per check, per tool, at scale.

Why this isn’t happening in isolation: the Google vs. SerpAPI case

This move sits inside a broader pattern. Google is actively suing SerpAPI under the DMCA for scraping search results via its anti-scraping system, SearchGuard. A federal judge dismissed Google’s original complaint in July 2026, finding insufficient evidence that copyright owners had actually authorised SearchGuard’s deployment. Google responded immediately, amending its filing on 10 August with specific licensing language from contracts with Reddit and image providers, explicitly restricting third-party extraction and resale of licensed content.

The pattern is deliberate: Google is pursuing a legal precedent and building technical barriers simultaneously. The /goto redirect is the technical prong. The lawsuit is the legal one. SerpAPI is the named defendant, but the infrastructure and the precedent affect every tool that depends on Google SERP data, which covers most of the rank tracking and SERP intelligence market.

It’s also worth noting that Google removed the &num=100 parameter in September 2025, forcing tools that retrieved 100 results in a single request to make ten requests instead. The goto redirect follows directly from that pattern.

For agencies, this has direct consequences. Rank tracker reports that were accurate last month may now be incomplete, with no error flag to signal the gap. The weekly keyword data that goes straight into client reports could be based on partial SERP reads if your tool hasn’t fully adapted. And the 500–1,000 request overhead creates a real cost decision for providers: charge more, check less frequently, or reduce keyword coverage. Over time, possibly all three. 

How the major tools are responding – as of 1 September 2026

The range of responses from rank tracking platforms tells its own story.

SE Ranking moved fast. Within days of the rollout, they posted publicly on LinkedIn: “Google’s new /goto links won’t affect SE Ranking’s rank tracking data.” Their engineering approach resolves the destination URL without touching the redirect, generating no artificial clicks in the process. Users see no data disruption and need take no action.

Ahrefs took longer, and their statement, when it came, was more candid. The official account posted: “This change has caused a temporary inconsistency in our data. When you’re processing the volume we do, these things take a moment to work through.” They cited their own web index and Yep.com as first-party data sources they can draw on if the pattern becomes entrenched.

The response drew significant pushback in the comments. One user pointed out that Ahrefs still hadn’t fully resolved data gaps from the &num=100 removal, and that Yep.com data is not a substitute for Google SERP data for paying customers.

Semrush posted on LinkedIn on 1 September confirming no impact: “Semrush has not been affected. Our team was aware of the change and addressed it quickly, so there’s been no impact on rank tracking, reporting, or data accuracy.” Their framing is notably direct: they describe the change as Google protecting its data from third-party tools, scrapers, and AI engines, and make clear they had adapted before it became a problem for users.

SerpAPI published a detailed blog post acknowledging impact across multiple products: their Google Search API, News Light API, Videos API, and AI Overview API. Encoded /goto URLs were returning in place of destination URLs in some cases. Their engineering team was working through variations as Google rolled them out, and they asked customers to report unresolved instances to speed up the fix. Given that SerpAPI is simultaneously defending a lawsuit from Google over exactly this kind of access, the measured tone is worth noting.

Nozzle raised the alarm earliest and with the most technical specificity. The 500–1,000 request figure came from their team, and they chose to discard affected result pages during testing rather than let incomplete data pollute their outputs.

The reporting risk agencies haven’t been warned about

The rank tracking story has been covered. What’s less discussed is the attribution risk sitting quietly underneath it.

The /goto redirect inserts an additional hop between the Google SERP and the destination page. When a redirect strips the HTTP referrer header (which can happen, depending on how the redirect is implemented), GA4 has no organic signal to read. It reclassifies those sessions as (direct)/(none).

The result looks like this: your GA4 organic traffic drops, your direct traffic ticks up, and nothing in the platform flags an error. It looks like a trend.

Meanwhile, Google Search Console continues reporting clicks accurately because it operates at the server level, not via referrer. So GSC shows steady or growing clicks while GA4 organic falls, a divergence that becomes increasingly difficult to explain to clients, especially if the baseline was already tight.

We cannot confirm widespread cases of GA4 misclassification yet. But the risk is real and the conditions are in place. The question is whether Google’s implementation strips referrers consistently or only in certain cases.

What agencies should do right now:

• Establish a baseline. Pull your GSC clicks vs GA4 organic sessions ratio for August, before full rollout. Set this as a reference point and flag if it shifts in September.
• Check your server logs. Look for google.com/goto appearing in referrer headers. If you see it, you’re already in a position to understand how it’s being handled on your setup.
• Know your GA4 referral exclusion settings. If misclassification confirms at scale, google.com may need to be added to the exclusion list. Know where that lever is before you need it.
• Get ahead of client conversations. If organic traffic appears to dip in August–September, you want to be the one framing it as a data anomaly, not the client asking why the numbers don’t add up.
• Ask your rank tracker provider directly. Don’t assume no news means no impact. SE Ranking has confirmed they’re unaffected. Ahrefs has confirmed a temporary inconsistency. If you’re using Semrush or another tool with no public statement, ask your account contact what their status is.

What this signals

Google has tightened access to its search data infrastructure three times in rapid succession: the &num=100 removal in September 2025, the SerpAPI lawsuit filed and amended in 2026, and now the /goto redirect at near-full rollout. Each step makes SERP data more expensive to retrieve and harder to access without Google’s cooperation. The legal and technical pressure is pointing the same way.

Rank tracking tools will adapt, as they always have. But adaptation has costs. Those 500-1,000 requests per query translate into higher operational costs for rank tracking providers, which translates into either slower refresh rates, reduced keyword coverage, or higher prices for users. Possibly all three over time.

The agencies best placed to navigate this are those who don’t treat third-party SERP scrapers as their only data layer. GSC integration, first-party analytics, server-level measurement, and a clear understanding of where each data source starts and stops, all matter more than they used to. In an environment where Google controls the pipes, reporting that depends entirely on third-party access to those pipes gets more fragile with each restriction.

For SUSO clients and partner agencies, your reporting isn’t dependent on any single data source. Our methodology layers GSC and GA4 data alongside position tracking as standard, in addition to AI search visibility tracking, so a disruption or data gap at the tool level doesn’t compromise the full picture. We’re monitoring the situation across all accounts and will communicate directly with clients and partners if anything changes.

FAQs

• Does this affect my website’s rankings?

  No. The /goto redirect changes how Google delivers links in search results, not how it evaluates or ranks pages. Your position in search results is unaffected. What may be affected is how accurately third-party tools measure and report those positions.

• Which rank trackers are confirmed unaffected?

  SE Ranking and Semrush have both confirmed their position tracking data is unaffected. Ahrefs has acknowledged a temporary data inconsistency and says their team is resolving it. SerpAPI has confirmed impact across several of their API products, with fixes in progress. If you use another tool, check with your provider directly rather than assuming silence means safety.

• Should I switch rank trackers because of this?

  Not based on this alone. The situation is still developing, and most tools will adapt – they always have when Google changes its infrastructure. A tool that is transparent about temporary data issues is arguably more trustworthy than one that says nothing. What matters is whether your provider communicates clearly and resolves issues promptly. Monitor that over the coming weeks before making any decisions.

• Is Google trying to shut down third-party SEO tools?

  Not explicitly, but the direction of travel is clear. Between removing the &num=100 parameter in September 2025, pursuing the SerpAPI lawsuit, and now the /goto rollout, Google is systematically making it harder and more expensive to extract data from its search results at scale. Whether that’s primarily about anti-bot measures, legal positioning, or data control is debatable. The practical effect for the tools industry is the same regardless.