The Fine Print Just Got More Expensive
Google Quietly Reroutes Search Links and Third-Party Tracking Could Take the Hit
New goto redirects and a possible human-verification step for search results threaten to scramble keyword tracking and AI citation data alike.
What's the News:
Google has confirmed it’s deploying goto URL redirects on search results links, swapping a visible destination domain for a google.com/goto?url string instead of the actual page URL. At the same time, Google is testing a “Verifying Your Request To Continue To Search Results” step that appears to require human verification before some searches load.
Why It Matters:
Both changes hit measurement, not just display. Replacing visible URLs with goto redirects could break how third-party rank-tracking platforms identify which domain is actually ranking, since most of these tools read the URL directly off the results page. And if the verification step rolls out further, it could block the automated crawlers that many non-Google LLMs rely on to source and train on Google’s results. This would be a direct threat to those models’ ability to cite web content accurately.
This is unlikely to cause anything more than short-term panic. Third-party platforms and non-Google LLMs will likely find workarounds, but expect real measurement disruption in the meantime.
Marketers Should:
Expect keyword ranking reports from third-party platforms to get noisier, or show a sudden drop-off, as goto redirects obscure destination URLs.
Watch citation and mention rates in non-Google LLMs for volatility as the verification step expands.
Use this moment to revisit your organic search measurement framework rather than waiting for the disruption to force your hand.
Build a system where each metric feeds the next one, with the throughline always tied back to business growth. Not platform-reported vanity numbers.
AI Writes Bad SEO Pages and Cannibalizes Its Own Rankings
A Search Engine Land case study shows what happens when Claude ships pages without a human check, and why most SEO teams are already exposed.
What's the News:
Search Engine Land contributor Will Scott documented two real cases where Claude, tasked with building new SEO pages, instead cloned existing homepages, swapped the title and H1, and reused the body copy. The result: zero impressions and zero clicks across six months of Search Console data, plus cannibalization of the original page’s rankings.
Why It Matters:
AI adoption in SEO is now the norm, not the exception: Keyword.com’s 2026 State of AI in SEO survey found 87% of SEO professionals use AI regularly, with 78% using Claude, specifically. That means the failure mode Scott describes isn’t a fringe risk, it’s sitting in most teams’ workflows right now. A cloned page that quietly cannibalizes an existing ranking doesn’t throw an error, it just underperforms. And by the time someone spots it in a monthly report, the client has already lost months of visibility and traffic that are hard to explain away in a QBR.
Never let AI push a page live without a human diffing it against what already exists on the site. That single check would have caught both failures in the case example. AI optimizes for a plausible-looking answer, not a right one, especially under time pressure, and “looks finished” is not the same thing as “is correct.”
Marketers Should:
Never let AI push a page live without a human diffing it against what already exists on the site.
Use Claude for what it’s actually good at: keyword clustering, content gap analysis, technical audit summaries, and first-draft outlines. Let the humans do the publishing.
Pull your last two months of AI-assisted pages and check impressions in Search Console; a zero-impression page is your canary for a silent duplicate.
Build the review gate into your process now, not after a client asks why a new page cannibalized an old one. If you can’t point to who signed off on a live AI-generated page, you don’t have a process yet.
The FTC Says Amazon Hid the Real Price of Its Ad Auctions
A lawsuit joined by 22 states alleges Amazon overrode its own ad auctions for years — and it's a warning for every marketer leaning on closed-platform metrics.
What's the News:
The FTC and 22 state attorneys general sued Amazon, alleging the company spent seven years secretly overriding its own advertising auctions — charging Sponsored Products advertisers their full winning bid nearly 80% of the time while continuing to describe the mechanism as a second-price auction. Amazon denies the allegations, arguing the lawsuit misunderstands how advertisers use its platform and that average costs stayed flat in real terms.
Why It Matters:
This is bigger than Amazon. Marketers are investing more into closed platforms where auction mechanics, pricing logic, placements, and optimization are largely controlled by the platform itself. As AI and automation take on more campaign decisions, advertisers need to understand whether growth is coming from true efficiency, higher prices, easy-to-capture demand, or mechanics they cannot fully see.
This lawsuit is a reminder that platform trust has to be earned continuously. The dashboard may show strong return on ad spend, cost per lead, or revenue…but those numbers don’t always explain how the auction worked, whether pricing changed, or whether the next dollar is actually incremental. The brands that win will use automation where it drives results, but pair it with stronger measurement, cleaner reporting, and better questions about incrementality.
Marketers Should:
Pressure-test platform-reported performance. Don’t stop at return on ad spend or cost per lead as reported by the platforms. Instead, use incrementality measurement tools to cross-check what the platforms are saying.
Watch for rising costs without better outcomes. If costs increase but order quality, lead quality, or revenue doesn’t improve, dig deeper.
Separate efficiency from incrementality. Low-cost conversions aren’t always new conversions. Make sure your media is creating demand, not just harvesting it.
Use contribution measurement where possible. Media mix modeling, incrementality testing, holdouts, and clean revenue reporting can show what’s really driving growth.
Keep humans close to automated systems. Automation optimizes inside the rules it’s given. Marketers still need to question the rules and audit the outputs.
We’re reinventing performance marketing because everything changed while you were reading this.
We help you navigate this new landscape with data-driven insights, platform expertise, and creative that connects, so your brand not only keeps up, but gets ahead.
Work with Us