The short version
- Google confirmed its
google.com/gotoredirect rollout in August. It has not publicly explained the October 30 verification behavior. - Serper says it now resolves the first three organic links immediately and offers a separate resolver for results four through ten.
- SerpApi lists an open incident for
/gotolinks appearing in organic result link fields. - Audit the exact destination URLs in your third-party data before interpreting an apparent ranking loss.
What changed on October 30?
In August, Google began putting an opaque google.com/goto?url=… address between some search results and their destination pages. Google described the rollout as a technical measure against abuse, according to Search Engine Land's report of its statement. A click ordinarily passed through that address to the publisher. Tools reading the search results could no longer extract the destination from the result link alone.
October brought a second complication. PrismCrawl reports observing a “Verifying your request” page after clicking some result links on October 30. Serper says Google now limits how many result links can be opened shortly after a search; in its response, it returns direct URLs for the top three organic results and /goto links for results four through ten, which it says work about 30 seconds later. Those are provider observations, not a Google announcement about a global rollout.
That timing matters. The search results page may load successfully while the result's destination is still unavailable to an automated reader. A successful API response can therefore contain a link that is unusable for domain matching, citation, or downstream crawling.
What is confirmed, and by whom?
| Observation | Source | Limit |
|---|---|---|
SerpApi's Google Search API returns some /goto links in organic_results.link. | SerpApi status and public issue | An incident at one vendor does not measure all search traffic. |
| Serper changed its handling: top three direct, later results through a resolver. | Serper status | Its stated 30-second window is vendor-specific. |
| Some human clicks showed an intermediate verification page. | PrismCrawl observation | It did not establish how widespread the behavior is. |
As of October 31, SerpApi's incident remains marked “Investigating”; Serper marks its incident “Resolved” after changing its response and adding /resolve. That difference is a reminder to check your own provider and query set rather than assume every tool behaves the same way.
What can break in an SEO workflow?
A rank tracker needs the destination URL to decide which page occupies a position. If a parser stores google.com/goto as though it were that URL, it may attribute the result to Google, fail to match your domain, or drop the row. Citation monitors and AI research pipelines face the same problem when they try to open or deduplicate a source.
This does not show that Google changed the ranking order or your indexing status. Search Console measures Search clicks and impressions within Google's own reporting systems. Its data is a useful cross-check, though it has different definitions and may lag third-party tools. A temporary mismatch between a vendor's rank chart and your Search Console trends is a reason to inspect the feed, not to rewrite pages.
For readers, the observed verification page can add a delay before the publisher receives the visit. We have no reliable evidence yet of a broad effect on traffic attribution or click-through rates.
A 15-minute audit before you act
- Export a small sample of raw organic results from October 29 to 31, including rank, link, displayed domain, query, location, device, and timestamp.
- Count links whose host is
google.comand path is/gotoor/url. Keep the count by rank bucket; the top three may behave differently from positions four to ten. - Check whether the provider supplies a separate resolved destination field or resolver. Do not treat a displayed breadcrumb as an exact page URL.
- Compare affected queries with Search Console clicks, impressions, and average position over a wider window. Segment by page and country before calling it a sitewide loss.
- Annotate the provider incident on your dashboard. Ask the vendor how unresolved links are represented, whether old data will be repaired, and whether retries change cost or latency.
If your raw links are clean and first-party performance also fell, investigate the usual SEO causes. This incident should not become a blanket explanation for every October decline.
What remains unknown
Google has not publicly described the October 30 behavior, its purpose, geographic reach, or duration. Provider tests differ by IP, session, result type, and time, so no single vendor status page can establish a universal failure rate. The current evidence also does not establish a causal link between these redirect delays and changes in organic rankings.
Our August /goto explainer covers the original redirect rollout and why third-party rank collection became more expensive. This update is about the new verification and resolution delays reported on October 30.