I used to write "what is" articles. Lots of them. "What is content marketing." "What is a backlink." "What is domain authority." They ranked, they drove traffic, they generated leads. That was the playbook for the better part of a decade and it worked brilliantly.
It does not work anymore. And I am not being dramatic. The data is ugly.
A recent analysis found that 73% of B2B websites experienced significant traffic loss between 2024 and 2025, with the average decline hitting 34% year over year. That is not a dip. That is a third of your traffic gone in twelve months. And the cause is not some mysterious algorithm update that you can recover from by tweaking your content. The cause is structural. The search engines are answering informational queries themselves, and they are getting better at it every month.
The numbers behind the collapse
Let me lay out the data because I think a lot of people are still in denial about how bad this is.
AI Overviews now appear on 99.9% of informational keywords. Read that again. Essentially every time someone searches for an informational query on Google, they see an AI-generated answer at the top of the page before they see any organic results. According to Ahrefs, the presence of an AI Overview correlates with a 34.5% lower average click-through rate for the top-ranking page. More recent data from 2026 suggests the impact is even steeper — organic CTR drops from 1.41% to 0.64% year-over-year when AI Overviews appear, which is a 55% reduction.
Zero-click searches now account for 69% of all queries. That means for more than two thirds of everything people search for, nobody clicks on anything. The answer is right there in the search results — sometimes in an AI Overview, sometimes in a featured snippet, sometimes in a knowledge panel. The user gets what they need and moves on.
Google's AI-powered search features now appear in 67% of B2B-related queries specifically. When they do, click-through rates drop to just 8%, compared to 15% for traditional results without AI features. And the trend line is unambiguous — CTRs for high-funnel informational queries are projected to be 20-30% lower than current levels by end of 2026, and there are no signs of recovery.
The bottom line is that if your content strategy depends on people clicking through from Google to read your "what is X" or "how to Y" articles, you are building on a foundation that is actively crumbling.
Why this is happening now
The timing is not random. Several forces converged over the past 18 months to accelerate the decline of informational content traffic.
Google expanded AI Overviews aggressively throughout 2025, moving from a limited rollout to near-universal presence on informational queries. The quality of these overviews improved dramatically as the underlying models got better. In early 2024, AI Overviews were frequently wrong or incomplete enough that users still clicked through to verify information. By March 2026, they are right often enough that most users trust them for straightforward informational queries.
ChatGPT, Perplexity, and other AI search tools absorbed a significant portion of informational search volume that previously went to Google. When someone wants to know "what is serverless computing" or "how does compound interest work," they are increasingly asking an AI chatbot instead of googling it. That traffic never even enters the traditional search funnel, so it does not show up as a decline in your Google impressions — it just vanishes from your analytics entirely.
And the content market saturated. After years of every company with a blog publishing "ultimate guides" and "complete overviews" of every topic in their space, there is simply too much informational content chasing too few remaining clicks. Google has more options than ever to source its AI Overviews, and individual publishers have less differentiation than ever to justify a click.
The content types that are dying fastest
Not all informational content is declining at the same rate. Some categories are getting hit harder than others, and understanding the pattern helps clarify what to do about it.
Definition content is almost completely dead as a traffic driver. "What is X" queries are now answered perfectly by AI Overviews in virtually every case. There is no reason to click through to an article that will tell you the same thing the AI just told you, but with more banner ads and cookie consent popups. If you have definition pages, they may still rank, but they are not going to drive meaningful traffic.
Basic how-to content is following the same trajectory, just a step behind. "How to calculate ROI" or "how to write a meta description" — these queries get clear, actionable AI-generated answers that satisfy most users. The traffic decline for how-to content is not as severe as for definitions, partly because some how-to queries are complex enough that the AI answer is insufficient. But simple, procedural how-tos are toast.
Listicle content — "10 best project management tools" or "7 ways to improve your website speed" — is holding up slightly better because users still want to browse options and compare, which is harder for an AI Overview to fully satisfy. But even here, the CTR decline is significant because Google often surfaces enough list items in the overview to satisfy casual searchers.
Overview and survey content — the classic "complete guide to X" format — is taking heavy losses. These were always written more for search engines than for humans, and now the search engines have decided they can provide the overview themselves. The era of ranking for broad informational terms by publishing a 5,000-word overview is effectively over.
What is actually growing
Here is where things get more interesting and more hopeful. While informational content is declining, certain content types are holding steady or actually gaining traffic. Understanding why reveals the new playbook.
Original research and proprietary data
Content that contains data nobody else has is the single strongest category right now. AI Overviews cannot cite data that does not exist elsewhere on the web. When you publish original survey results, proprietary benchmarks, or unique datasets, you become a primary source that models need to reference. You are not competing with the AI answer — you are feeding it.
I have seen this firsthand with client content. A B2B SaaS company I work with published a quarterly benchmark report based on anonymized data from their platform. That report generates more traffic now than it did a year ago, even as their informational blog content declined 30%. The reason is simple — journalists, analysts, bloggers, and AI models all need to cite original data, and there is no substitute for it.
The investment required is real. Running surveys, building data pipelines, analyzing datasets — none of that is cheap or fast. But the content assets it produces have a durability that "what is" articles never had.
Experience-driven content
Content that describes genuine first-hand experience with a topic — not theoretical knowledge, but practical, opinionated, been-there-done-that accounts — is performing well because it is something AI cannot replicate. An AI can tell you what the best practices are for migrating a website. It cannot tell you what it felt like when the migration went wrong at 2 AM and how you fixed it.
Google has been pushing the E-E-A-T framework (Experience, Expertise, Authoritativeness, Trust) for years, and the "Experience" signal is now doing real work in distinguishing human content from AI-generated overviews. Content that says "I tested this and here is what happened" performs differently in the ranking algorithms than content that says "experts recommend the following."
This is one area where smaller publishers and individual practitioners have an advantage over large content operations. A single SEO consultant writing about their personal experience optimizing a real website will outperform a content team producing generic advice based on secondhand knowledge. The authenticity gap is widening and search engines are getting better at detecting which side you are on.
Community and discussion content
Reddit's rise to the third most visible domain in Google search results is not an accident. People increasingly want to hear from other people, not from brands. They want unfiltered opinions, real experiences, and the kind of messy, imperfect information that only comes from community discussion.
This is why forum content, community posts, and user-generated discussion threads are gaining traffic while polished brand content is losing it. The implication for content strategy is that creating spaces for your audience to talk — and then making that discussion accessible to search engines — may generate more organic traffic than publishing more articles.
Tool and interactive content
Calculators, templates, assessments, configurators, and other interactive content are holding up well because they provide utility that a text-based AI answer cannot. If someone wants to calculate their mortgage payment, an AI can explain the formula, but a mortgage calculator lets them actually do it with their specific numbers. That functional value drives clicks even when informational value does not.
I have been advising clients to convert their best informational content into interactive tools wherever possible. A "how to calculate customer acquisition cost" article becomes a CAC calculator. A "how to choose the right project management methodology" article becomes a recommendation quiz. The content still targets the same search intent but it delivers value that an AI Overview cannot replicate.
What to do about it
If you are sitting on a content strategy that is primarily informational — and let's be honest, most B2B content strategies are — the instinct might be to panic. Don't. But do move.
Start by auditing your existing content for the types I described above. Categorize each piece as definition, how-to, listicle, overview, original research, experience-driven, community, or interactive. Calculate the traffic trend for each category over the past 12 months. You will almost certainly see the pattern I described: informational categories declining, experiential and data-driven categories holding steady or growing.
Then shift your content production calendar. This does not mean abandoning informational content entirely — some of it still serves conversion purposes even with reduced traffic. But the ratio should change. If you were spending 70% of your content budget on informational pieces and 30% on everything else, flip it. Invest in original research, first-person case studies, interactive tools, and community-driven content.
For the informational content you keep producing, optimize it for citation rather than clicks. Accept that many users will never visit your page and instead make sure the information on your page is the information that AI Overviews cite. Include specific statistics, clear definitions, and authoritative statements that models are likely to extract. This does not drive traffic directly but it drives brand visibility, which has its own long-term value.
And seriously consider whether some of your content budget should go to other channels entirely. YouTube, podcasts, newsletters, and community platforms are not subject to the same AI Overview dynamic that is crushing web traffic. A dollar spent on a YouTube video or a newsletter might generate more audience engagement than a dollar spent on a blog post that Google will summarize before anyone clicks on it.
The uncomfortable truth
I want to be honest about something that the content marketing industry does not like to hear. The golden age of informational SEO content — where you could publish well-optimized articles about common questions and watch the traffic pour in — lasted about ten years, from roughly 2013 to 2023. It was a beautiful era and I built a career during it. But it is over.
What is replacing it is harder. Original research costs more than desk research. Experience-driven content requires having actual experience. Interactive tools require development resources. Community building requires patience and authenticity that cannot be faked.
But that difficulty is also the moat. If creating valuable content were easy, everyone would do it and it would stop working, which is exactly what happened with informational content. The content types that work now are harder to produce, which means there is less competition and more reward for the companies willing to invest in them.
The informational content era rewarded volume. The post-informational era rewards depth, originality, and genuine utility. I know which one I would rather compete in.