Data gathered by Chartbeat and shared by Axios reveals that, over the past year, Google Search traffic to publishers across the broader web have fallen drastically, and proportionally more so for smaller websites. Referral traffic from Google apparently fell by 60% for “small publishers,” while “medium publishers” (those with between 10,000-100,000 daily pageviews) saw a drop of 47%. “Large publishers,” meanwhile, saw a 22% drop. That last category would be any site getting over 100,000 daily pageviews.
It’s not just Google Search either. While Search traffic dropped by 34%, traffic from Google Discover has also fallen by 15% over the past year, the report found.
No, fewer people getting past the AI summary
Holdup, are people not skipping the AI summaries entirely because the info is fucking shit?
You forget we are in an echo chamber here. Most people not only read the AI summaries, they believe them. Just the other day I saw a normie ask ChatGPT to add up some numbers for them, instead of using a calculator. That’s how entrenched AI has become in their day-to-day. They don’t have to think any more. Thinking is hard. And that’s how Google is able to dominate the web. Steal the data and serve it up as slop that’s good enough for the everyday Joe.
Given the state of a lot of the summaries I’ve seen lately, that is scary.
Is that what this is saying? I wasn’t sure. The article should state that explicitly, and not assume that the reader concludes that.
Hard to imagine usage of Google suddenly falling by 22%, much less 60%.
Good news, though, is if Google stops bringing in traffic to sites, they’ll block its bots, so both search and Gemini will become even worse, possibly turning people away.
I think the issue there is the data doesn’t tell anyone “why”, it only tells “what”.
This is actually a good thing. Google get paid for referrals and niw their “AI” shit turns against it.
What are you talking about?
The only use I have found for the AI summary is quickly getting NAIC numbers for insurance companies at work. Otherwise I use an extension that removes the AI summary.
Are those results correct, though?
They actually are surprisingly.
Which is probably enough to find the info 90% of the time
I have classic apple computers.
I also maintain a small list of sites I visit to get abandonware programs for them. Of the times I’ve used the AI results, I found what I was looking for fewer than 15%. At one point, I had the AI telling me there was no such thing as Winamp for Mac, while I was running it in MacOS 8.6 under the virtualization program, Sheepshaver.
Seriously?
AI’s got so little ability to sort through archived knowledge and pull up old links and sources, it’s as if anything before 2006 never existed.
Nuts to that.
I hit up ten blue links and have never looked back.
But did a regular search provide the correct info? I find niche searches aren’t always good using either method. Old software info can be hard to find.
Yes! I wasn’t looking for whether it existed, I knew it did, but it was in a .sit file with an abbreviated name. Also apparently was an aplha build, so maybe that’s why the AI insisted it did not exist. Was looking for the last version available for the classic OS as I had one of the earliest.
deleted by creator
this is problematic on multiple levels.
Name 3 levels
It helps spread false information widely
It puts a lot of control of information in a single companies hands
It hurts the underlying sources
When google provides the info directly, and the first hand sources has become completely obsolete and shut down, what would new information stem from? It’s an inherently unstable and short sighted solution.
Google controls search results and has been caught meddling. Which negates the first two.
The last one of hurts the sources… sure they get less traffic which is less ad revenue. Cry me a river.
Level one, level two, level three. WHAT NOW, BITCHES?