Updates by Casey Anderson, Paid Media Director
As AI-driven, zero-click search keeps cutting into organic referrals, publishers are paying to recover the visitors they used to get for free. According to Similarweb data reported by Adweek, the Top 100 media publishers spent an estimated $113 million on paid search in July 2026, up 41% year over year and 274% over three years. Forbes alone spent about $72 million that month, while its organic search traffic fell almost 27%, and CNN and USA Today saw similar organic declines.
The spending isn’t a blanket fix for lost traffic. Publishers are putting their money where the return is easy to measure, mainly commerce and product content, where a single affiliate click can be worth around $20, far more than a standard display ad. That’s why the most aggressive spenders bid on high impact terms like pet insurance, high-yield savings accounts, and GLP-1 medications, sometimes paying up to $50 per click, while most keywords cost $1 to $3. Some analysts see this as smart resource allocation: publishers are doubling down on what works.
The strategy has a catch, though. Publishers competing for the same keywords drive up the price of the traffic they’re all trying to buy, and every dollar goes to Google, whose AI search products are shrinking their organic reach to begin with. For the industry, paid search looks less like a long-term solution and more like a costly bridge while publishers build more durable channels, such as apps and direct audience relationships.
Since U.S. testing began in February, ChatGPT ads have expanded to more than 60 countries. OpenAI also rolled out updates to make campaigns easier to scale and measure.
Advertisers can now create product feed campaigns in bulk through spreadsheet uploads, see why individual products were rejected in review, and track attributed conversions beyond a campaign’s main optimization goal. OpenAI says product-feed ads have been among its strongest performers.
Early benchmarks are starting to emerge. A Graphite report using Similarweb data found click-through rates of 0.6% to 0.75% in the U.S., U.K., Canada, Australia, and New Zealand, compared with more than 1% in Brazil, Mexico, Japan, and South Korea. Because the report pools clicks across all advertisers, these figures are rough guides rather than direct comparisons for individual campaigns.
Advertisers are still experimenting. Only two brands held a top-10 spot every week from May through mid-September, and B2B SaaS led U.S. impressions while retail lagged. With bulk product tools, a Shopify partnership, and an Amazon Ads pilot, OpenAI is clearly courting retail spend, so now is a good time for brands to move past small tests and embrace scaling while it’s still early.
Google keeps expanding AI Max for Search, which lets its AI find relevant queries beyond an advertiser’s keyword list, using landing pages and other campaign signals. As adoption grows, Google is adding ways for advertisers to guide the automation and see what it’s actually doing.
Google is also building a unified report that connects the search term that triggered an ad, the creative shown, and the landing page reached, all in one view. This will all be through AI Brief, which allows brands to provide AI Max with written context about their business, audience, and key messages. There’s no launch date yet, and more details are expected later in 2026.
Until that arrives, advertisers can already measure how much AI Max contributes to results. The Search Terms report often groups a large share of activity under “Other search terms,” so it can’t show the full picture. The Keywords tab can. Its summary rows separate AI Max expanded matches and landing page matches from your own keywords, and adding those two together and dividing by total campaign conversions gives you AI Max’s share of conversions.
Tracking that share weekly by campaign helps teams focus their attention. When AI Max’s contribution jumps in a campaign, that’s the place to dig into search terms and check whether the new traffic is relevant, converting at an acceptable cost, or getting too broad or too branded. Google is pairing more automation with more transparency, and advertisers who track AI Max closely will be better positioned to use it well.
→ Need help getting better performance from your Google AI Max campaign? Our Paid Media Audit provides an expert view.
Head to our October Analytics Roundup for more details on updates to Google Ads’ measurement suite including Data Manager and Meridian GeoX.
Updates by Brett Woodward, Sr. SEO
In September, Google announced it was rolling out its fourth spam update of the year, which is the highest number of spam updates we’ve seen. If that doesn’t just sum up the state of search and the internet in general these days, I don’t know what does.
AI is great at helping us do things at scale and programmatic content does have its place in certain situations, but low-quality content has become incredibly pervasive and Google is taking notice.
If you’re using AI to make content, you need to ask yourself if the content you’re creating has unique value for real users or if you’re doing it for keyword and topic coverage to bolster your site’s authority. If it’s the latter, you might want to reconsider your strategy before you’re hit with a manual action from Google, which will no doubt affect both your search and AI visibility.
→ Watch Brett’s video on building content to serve your audiences.
A new filter in Google Search Console allows you to filter results for multimodal searches. Multimodal refers to a search done with the help of an image, like when you upload a photo to Google Lens.
Results using this filter won’t give you query data, but you can see which pages on your website are being shown or ranked for image search.
This information can give you an idea of what kinds of images are getting visibility on search engines and can provide you with some clues as to what kind of content to produce more of in the future.
Over the past few months, several SEOs like Chris Long have been monitoring the increase of AI overviews being included on searches for brand names. But over the course of the last month, new data from DemandSphere indicates that the instances of AIOs for branded queries have jumped from 25% to over 80%.
Since simple branded searches usually have navigational intent, and in my opinion it doesn’t make a lot of sense to throw up an AIO, but Google’s gonna do what Google’s gonna do. This isn’t happening for smaller companies from what we can see, as the AIOs were found mostly on branded searches for large corporations.
The fact of the matter is that big companies are now going to need to deal with LLM interpretations of their brand being put in front of users who were just trying to get to their websites.
If you’re a user determined to get to a website, these AIOs probably won’t be a major deterrent, but I believe that this could become a serious hurdle for users who are just looking for information on a brand.
Rather than getting the information about a company directly from their website, users may get enough information on the search engine result page to deter them from reaching the website. Or deter them from engaging with that brand at all, for that matter.
Ann Smarty’s recent post on LinkedIn highlights the kind of friction these branded AIOs could cause. In her example, an AI Overview mentioning the hosting platform GoDaddy generated a section on “Why Many Avoid It”, which could seriously impact whether a user chooses to continue considering that company.
To be transparent, in the comments Ann mentioned she had been searching queries around reviews and complaints before making this unspecific search, so this response was likely triggered by her previous searches.
Regardless, what this shows us is a situation where the AI overview has the potential to shape the user’s perception of a brand before they can even visit the company’s website.
The takeaway here is that it has never been more important to control the narrative of your brand to the best of your ability. It’s also critical to review the sentiment of your brand online, improve your product or service, and earn the widespread trust of your customers. Because if you don’t, other people’s opinions may end up shaping the narrative around your brand before you can get a word in.
Head on over to the October Analytics Roundup, where this month Head of Analytics Nico Brooks shares product updates from Google Analytics, Google Tag Manager and Data Studio, as well as some great articles on AI-assisted analysis, data visualization and more.
The art of throwing things away; product updates in Google Ads, GA4, Data Studio, GTM;…
Learn where you can impact how your brand is described outside your website, so you…
Developing trustworthy architectures for insights with AI; Measurement Protocol replaced by Data Manager; and lots…