The Measurement Gap Is Widening: Google Breaks Rank Tracking, AI Mode Goes Dark, and ChatGPT Ads Hit $1B
Google's /goto Redirect Is Breaking Your SEO Reports, Not Your Rankings
Third-party keyword counts are collapsing while real traffic holds flat, and most executive dashboards haven't caught up.
What's the News:
In August 2026, Google confirmed it is routing search result clicks through a google.com/goto redirect. It’s the third change since January 2025 aimed at blocking tools that scrape search results. In one example from the analysis, Semrush data showed Reddit losing more than 60 million tracked keywords (a 25% month-over-month decline) while its estimated organic traffic stayed flat.
Why It Matters:
The keyword counts and visibility scores many teams report on no longer match what’s actually happening in the business. The position data shows the pattern: tracked results in position 1 rose 16%, positions 11 to 20 fell 28%, and positions 51 and beyond fell 35%. That’s lost tool coverage, not lost rankings. If your monthly deck leads with keyword totals, you could be showing leadership a decline that never happened, or asking for budget to fix a problem that doesn’t exist. Year-over-year Search Console comparisons have also been skewed since the num=100 change last September, so October 2026 will be the first clean comparison month.
Marketers Should:
Remove total keyword counts and third-party visibility scores from the executive summary this month. They measure the tool’s coverage now, not your performance.
Re-anchor organic reporting on revenue, conversions and first-touch attribution. Those numbers held steady while keyword counts fell off a cliff, which is exactly why they belong at the top of the report.
Keep using rank tracking for positions 1 through 10. Beyond position 20, treat the numbers as directional at best.
Mark October 2026 as your new year-over-year baseline in every dashboard. Comparisons that run through last September’s num=100 change aren’t comparing the same thing.
Expect Google to keep tightening access. Three changes in 20 months is a trend. Reporting built on your own analytics and CRM data won’t break the next time Google makes a move like this.
Google AI Mode Now Watches the Web on Your Customers' Behalf
Info monitoring is live for every user globally, and there's no way yet to see whether it's sending anyone to you.
What's the News:
Google is expanding information monitoring in AI Mode from Ultra and Pro subscribers to all users globally, announced by Robby Stein, VP of Product at Google Search. Users set up a standing query (tracking new restaurant openings nearby, back-in-stock alerts, price drops) and AI Mode continuously scans sites, forums, social posts and Google’s real-time data sources, including its 60-billion-product Shopping Graph, then sends updates when something changes. Google will also proactively suggest monitoring tasks while people search.
Why It Matters:
This turns AI Mode from a single-query tool into a standing surveillance layer between your brand and the customer. If a user sets up monitoring for “restock alerts for [product category]” or “new locations opening near me,” your visibility now depends on whether Google’s monitoring picks up your inventory changes, price drops or openings, not just whether you rank well on a one-time search. Local and shopping use cases are exactly what Google is leading with, so retailers and multi-location businesses are the most immediately exposed.
The open question is measurement, and Google hasn’t answered it. Nobody outside Google can currently see how often a monitoring task surfaces a given business, or whether those updates carry clickable links at all. Until Search Console or Merchant Center adds reporting tied to monitoring, this is an invisible channel: it could be generating traffic or awareness, but there’s no way to attribute it.
Marketers Should:
Run your own test monitoring tasks for the categories you compete in. See whether your pages actually surface, and whether those results link out.
Keep inventory, pricing and location data current in Merchant Center and Business Profile. Monitoring is only as good as the real-time signals Google can read, and stale data means you won’t trigger the alert.
Treat it the way you treat AI Overviews and Discover. Assume it’s already shaping demand, and track indirect signals like branded search lift and direct traffic since Google isn’t handing over the direct ones.
Watch Search Console and Merchant Center for monitoring-specific reporting. The moment it appears, add it to your baseline.
ChatGPT Ads Hit a $1B Run Rate. The Playbook Is Still Unwritten.
New research shows what early winners are doing differently, and why your Google and Meta creative won't transfer.
What's the News:
New research from Graphite and Similarweb, covering more than 98,000 ads and 8,000 landing pages, shows ChatGPT Ads have hit a $1 billion annualized revenue run rate, and 43% of ad-eligible U.S. users have seen an ad six months in. Click-through rates have doubled since launch, but the researchers still call it an experimental channel, not a core acquisition channel. Google still carries far more ad volume, so ChatGPT is framed as incremental, not a Search replacement.
Why It Matters:
The channel is developing its own behavior and creative rules fast enough that copying a Google or Meta playbook underperforms. Top advertisers write 98% new copy built around the actual use cases people bring to a conversation rather than repurposing existing ads. Most use five or fewer landing pages, reusing high-intent pages that already convert instead of building a new destination set. And timing within the conversation is now a real variable: ads in the first few turns get the strongest CTR and impression share, but a longer tail of opportunity exists after those early exchanges.
It’s still early enough to treat ChatGPT Ads as a test-and-learn channel rather than a shift in core strategy. Even Graphite recommends an experimental allocation over full commitment.
Marketers Should:
Carve out a small, dedicated test budget instead of pulling from Search. The data supports incremental, not replacement.
Build native creative from scratch. This channel can’t inherit whatever RSA or Meta assets are already in rotation. It needs copy written for how people actually talk to ChatGPT.
Start with your existing high-intent pages. Feature, pricing and product pages that already convert are the right destinations. Hold off on building ChatGPT-specific landing pages until the data justifies them.
Move now while competition is low, but frame results as directional. With no settled playbook, early performance shouldn’t be benchmarked against mature channels yet.
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