I used to tell clients that a good content strategy was patient work. Build the right pages, target the right keywords, wait for Google to reward you, and the traffic compounds for years. That advice held up for a long time. It doesn't anymore, and pretending otherwise is how you burn a marketing budget on pages nobody will ever click.
Something broke in the last two years, and it wasn't a small algorithm tweak. It was the entire premise that publishing answers to common questions gets you found. Google now answers those questions itself, right there on the results page, before anyone reaches your site. And the people who are winning right now aren't the ones with the biggest content libraries. They're the ones readers actually seek out by name.
Did You Know
Google search referral traffic to publishers fell roughly a third globally in the year to November 2025, with U.S. publishers hit even harder at around 38%, according to Chartbeat data cited in the Reuters Institute's Journalism, Media and Technology Trends report. If you're planning content the same way you did in 2022, you're planning for a search engine that no longer exists.
The Traffic Model We All Built On Just Stopped Working
Here's the part that should worry anyone still running a keyword-first content calendar. When Ahrefs compared click-through rates before and after AI Overviews became widespread, they found that pages ranking in position one for informational keywords saw their click-through rate collapse from 7.3% to just 1.6% once an AI Overview appeared on that query. That's not a dip. That's a page doing everything right and still getting almost nobody through the door.
The pattern shows up everywhere you look. Business Insider's organic search traffic fell 55% between April 2022 and April 2025, which is part of why the company cut over a fifth of its staff last year. Wired lost more than 60% of its Google traffic over a similar stretch, and some smaller sites dropped over 90%. None of these were bad publishers. They were running the exact playbook the industry taught for two decades, and the ground moved under them.
| Metric | Before AI Overviews | After AI Overviews |
|---|---|---|
| Organic CTR, informational query, position 1 | ~7.3% | ~1.6% |
| U.S. publisher Google referral traffic | Baseline | Down ~38% YoY |
| Google Discover referrals (2,500+ sites) | Baseline | Down ~21% YoY |
| Share of Google searches without any click | ~50% | ~60%, up to 77% on mobile |
What that table tells me isn't that content stopped mattering. It's that a specific kind of content stopped mattering, the kind built to answer a generic question that a machine can now answer faster than you can load your homepage. If a page can be fully replaced by a three-sentence summary, it was never really doing much for you in the first place. It just felt like it was, because for years that was the only game in town.
Where the Attention Actually Went
While publisher traffic was falling, something else was climbing steadily in the background: individual writers building audiences that don't route through Google at all. Substack now hosts nearly 100,000 monetizing publications, almost double the 50,000 counted a year earlier, and gross writer revenue on the platform crossed $450 million in 2025. That's not a niche hobby anymore. That's a real, growing economy running parallel to the one that just got gutted.
The names moving there tell their own story. Paul Krugman left the New York Times after 25 years to publish independently. Jim Acosta walked away from CNN and reportedly picked up over 10,000 paid subscribers within weeks. These aren't people chasing a trend. They're people who realized their name had become worth more than their masthead.
💡 The Reverse Halo Effect
For most of media history, working at a big-name outlet lent you credibility. Now it often runs the other way. A recognizable individual brings the audience, and the brand or platform is what benefits from being associated with them. Once you notice this pattern, you start seeing it everywhere, not just in journalism.
I don't think this means every brand needs to hand its content strategy to freelance personalities and hope for the best. But it does mean the old assumption, that a strong domain and enough published pages will earn trust on their own, doesn't hold the way it used to. Trust has started attaching to people again, the way it did before the internet made "content" a numbers game.
"Evergreen" Was Never the Content. It Was the Convenience.
I want to be careful here, because I've seen this get oversimplified into "evergreen is dead, stop writing guides." That's not quite right. What's dead is the version of evergreen that exists purely to catch search volume. A comparison page written so an AI can lift it into a summary. A "what is X" explainer with nothing in it that required actual work to produce.
What still holds up is content that requires you to have done something. Run the test, worked the client, seen the data nobody else has seen. That kind of writing doesn't compress into a three-line answer, because the value isn't the information, it's the fact that you're the one who has it. If you want a deeper look at how we think about separating information from insight when we plan client content, we wrote about it in our piece on why content fails to rank and what actually separates trustworthy work from noise.
✅ A Quick Gut Check Before You Publish
Ask yourself honestly: could a language model produce this exact page from public information alone, with no access to your data, your clients, or your specific experience? If the answer is yes, you're producing raw material for someone else's summary, not something a reader needs to visit you for.
Owned Distribution Is the Only Kind That Can't Be Taken From You
This is the part that matters most for brands, and it's easy to miss if you're only reading the headlines about journalists jumping ship. The reason the Substack migration and the collapse of evergreen are really the same story is distribution. Keyword-driven content came with a built-in audience, because Google handed you the reader. That relationship is thinning out fast, and it isn't coming back the way it was.
Newsletter subscribers, members, and people who type your name into a search bar instead of a topic are the only audience a platform change can't quietly take away from you overnight. We've talked before about why we push clients toward owned channels even when paid and organic are performing well, in our post on how to stay visible as search keeps evolving, and this is exactly the scenario that argument was built for.
| Distribution Type | Who Controls It | What Happens When the Platform Changes |
|---|---|---|
| Search ranking | Can vanish with one algorithm or AI Overview update | |
| Social following | Meta, X, LinkedIn, etc. | Reach throttled or deprioritized without notice |
| Email list / newsletter | You | Portable, exportable, and yours regardless of platform decisions |
| Direct / branded search ("your company name") | You, built over time | Grows as reputation grows, not as a rented placement |
The useful test I picked up from watching how larger publishers are adjusting their planning: would this piece of content still make sense to invest in if search sent you zero visitors tomorrow? If the honest answer is no, you're not building an asset, you're renting attention you don't control and calling it a strategy.
People Don't Prompt the Way They Search
There's a second habit that needs to go along with keyword-chasing, and it's how most teams still measure whether their content strategy is working. Ranking trackers assume someone types three words and scans ten blue links. That's not how people use AI tools. They describe a real situation, add constraints, ask follow-up questions, and land on a conclusion after a back-and-forth, not a single search.
If you're still measuring visibility purely by keyword position, you're measuring a behavior that's shrinking. The better approach, and the one we've moved toward with clients running content programs alongside us, is to build out realistic prompts pulled from actual customer language, the kind you hear on sales calls or read in reviews, and track where your brand shows up across AI platforms when someone asks those questions in their own words. It's more work than pulling a rank report. It's also the only version of that report that reflects how people are actually behaving.
Did You Know
Brands that get cited directly inside AI Overviews earn somewhere between 35% and 120% more organic clicks than uncited brands on the same queries, depending on the study. Getting mentioned by the AI isn't a consolation prize anymore. It's quickly becoming the actual prize.
What This Actually Looks Like Week to Week
None of this means throwing out a content calendar and hoping inspiration strikes. It means shifting what earns a slot on it. Here's how I've been reframing the conversation with teams working through this shift right now:
- 1Put a name and a face on your expertise. If a piece could have been written by anyone, it will get treated like it was written by no one, including by the AI systems deciding what to cite.
- 2Build the list before you need it. A newsletter with a few thousand engaged subscribers is worth more than a spike in organic traffic that could disappear with the next update.
- 3Do the thing before you write about it. Run the experiment, ship the product, talk to the customer. First-hand experience is the one input a summarizer can't fake.
- 4Track brand visibility across AI answers, not just Google rankings. Set up a way to see whether ChatGPT, Perplexity, and AI Overviews are mentioning you at all, because that's a different game than tracking position three on a keyword.
- 5Stop producing anything "just for SEO." If the only reason a piece exists is to catch search volume, it's the exact kind of content already being replaced by a two-sentence summary.
We put a longer breakdown of how we're restructuring editorial calendars around this shift, including how we prioritize which topics still deserve a full article versus a shorter answer, in our recent post on content optimization techniques built for how AI search actually works.
Where This Leaves Marketers Like Us
I don't think brands are powerless here, whatever the more dramatic headlines suggest. Brands still have resources, distribution, and the ability to showcase expertise at a scale no single person can match on their own. What's changed is who gets the credit and the click when that expertise shows up. It goes to the name attached to it, not the domain hosting it.
The businesses that come out ahead in the next few years won't be the ones who published the most pages. They'll be the ones who gave their expertise an actual voice, built a way to reach their audience directly, and stopped mistaking search rankings for a relationship. That's a harder thing to build than a content calendar full of keyword targets. It's also the only version of this that a platform update can't take away from you overnight.
