I took some grim satisfaction in Google’s very bad Thursday, a consequence of the same forces that the company is using to deprive news organizations of click-throughs to their websites. (And no, I will not refer to the company as Alphabet, although it is in the soup.)
Sign up for free email delivery of Media Nation — and become a supporter for just $6 a month.
Google’s stock price declined by nearly 13%, costing it $500 billion in market value. A major part of the reason is that those AI summaries Google is posting at the top of search queries are doing exactly the same to the tech giant as they are to media companies: keeping users on site rather than clicking through to a page that would serve up some ads.
Sara Fischer reported at Axios in March that “referral traffic from traditional search engines has declined by 60% for small publishers, compared with 47% for medium-sized publishers and 22% for large publishers, per Chartbeat.” In other words, the Google Zero phenomenon has hit local news outlets especially hard.
Now Google Zero has come … for Google. As Gizmodo put it in a headline, “Google Eats Its Own Tail (and Profits) With AI Summaries as Stock Price Plunges.” AJ Dellinger writes:
Philipp Schindler, chief business officer at Google, said that the AI-generated responses that Google now overwhelmingly serves to users whenever they search for anything have made it more difficult for Google to monetize results, cutting into its usually extremely profitable advertising business and contributing to the company missing its revenue projections in its latest earnings call.
And it’s not like Google can walk away from AI search. If it does, users will simply gravitate to another AI platform. My main search engine these days is Claude, although I hasten to add that I always ask for links to credible information.
Meanwhile, Reddit has announced that it’s moving away from the open web and instead preparing to require that users log in. The goal, writes Matt Growcoot of PetaPixel, is to make it harder for AI bots to scrape the site:
One of the main reasons behind the decision is the growing volume of automated activity targeting Reddit’s content. The platform contains a substantial collection of discussions and user-generated material, making it a valuable source of information for AI companies. Automated tools can access the website and collect large quantities of posts, often without the company’s permission.
As we wait for the AI bubble to burst, it’s fascinating to see how Big Tech’s best-laid plans are starting to run afoul of reality.
Discover more from Media Nation
Subscribe to get the latest posts sent to your email.
Thanks for this interesting update. Perhaps there will be more unintended negative consequences to befall the tech Icaruses in coming days/weeks/months/years…
I know I’m supposed to agree with all this as a hyperlocal news publisher. However, I think the system that sent so much referral traffic to all of us — and boosted Google’s profits — became an SEO scam geared at fooling people to click on stuff they don’t care about, rewarding a lobotomizing crazed competition for people’s momentary intention, and increasingly fake and meaningless metrics based on bot traffic (rather than human traffic). A system that 20 years ago benefited society (anyone in the world could find good reporting they were interested in from the smallest outlets) became a grotesque competition for eyeballs controlled by SEO-gaming corporations and scammers. If our reporting (as opposed to pink slime reporting) contributes to accurate Gemini searches that give people smart information (a big if), along with links for people truly interested in our full stories for more depth — then I think the public is better served. Of course the question is complicated, and much about AI is dangerous. But I disagree with reflexively bashing what might be an advance in the process of informing the public simply because it disrupts a bogus predatory business model that Google and major publishers came to rely on. I’m not crying for their profit losses and tanked Google Analytics metrics.
Paul, I’m going to push back. You’re the executive director of a nonprofit news organization, which means you have different incentives than a for-profit. You could make the argument that you’re making here to a prospective funder, for example. It’s kind of like Facebook: If a nonprofit has a lot of followers, well, that’s a story it can tell funders. For-profits are dependent on users clicking directly through to their site so they can sign them up as subscribers or sell them ads.
Some public TV stations are developing channels on YouTube. Again, that’s a story they can tell funders. For-profit TV newscasts need you to sit down and watch TV.
Where we might agree is that search is no substitute for having a direct relationship with your audience. Especially if you’re local, you should be building that newsletter network and not depending on people finding random stuff from your site on Google.