Most companies still treat their blog as an asset. Hundreds of posts, years of effort, a library that looks like proof of work. Open the analytics, though, and a lot of it does nothing. No clicks. No leads. No reason for anyone to find it.
This is not a traffic dip. It is a shift in what content is worth.
AI search changed the rules. AI Overviews, AI Mode, ChatGPT, Claude, and Perplexity now answer basic questions inside the search experience. Users get the definition or the quick tip without clicking anything. Nobody ever sees the generic post that used to catch that click.
So a meaningful portion of the average blog library is now functionally invisible. Not deleted. Just ignored.
What “content devaluation” actually means
Content devaluation is when a page stays indexed but stops earning anything. Google can still crawl it. A user could still land on it. But it pulls no weight.
You can spot it in the data. Watch for these signs:
- Impressions sliding down month over month
- Clicks near zero for queries that used to convert
- No leads or assisted conversions tied to the page
- No backlinks pointing in
- Heavy overlap with two or three other posts you published
- Nothing on the page that AI cannot summarize in a sentence
Invisible does not mean removed. It means the post no longer earns attention, clicks, or business value. The page is alive in the index and dead on the balance sheet.
How much of your library fits that description depends on the library. For some companies it is a slice. For others, an audit turns up 30%, 50%, or more of older content. None of it has a real job left.
The wider web shows the pattern at scale. Ahrefs studied billions of pages and found that more than 96% get no organic search traffic at all. Most published content was already quiet. AI search is now turning more of it silent.
Why AI search hit generic content hardest
Think about why someone clicked a “What is SEO?” post three years ago. They wanted a fast, clear answer. Now they get it inside the search result or from an assistant, no link required.
That single behavior change guts a whole category of content. Generic top-of-funnel posts existed to capture simple questions. AI answers those questions for free.
Pew Research put numbers on it. When an AI summary shows up, users click a normal search result in just 8% of searches. Without a summary, that figure is 15%. Users click links inside the summary itself about 1% of the time.
Look at the titles taking the hit:
- “What is SEO?”
- “What is branding?”
- “Benefits of content marketing”
- “5 tips for a better website”
These topics are not dead. But a basic explainer on any of them now needs a stronger reason to exist. If the page only restates common knowledge, an AI system has no reason to surface another copy of it. There is nothing on the page it could not generate itself.
The survivors carry something the model cannot fake: a real example, a sharp opinion, data you collected, a decision the reader actually faces.
Google’s real line: commodity vs non-commodity
Google’s guidance keeps circling one idea. It wants unique, helpful, people-first content. Strip away the phrasing and you get a simple split.
Google’s own guidance on helpful content asks two blunt questions of any page. Does it provide original information, reporting, research, or analysis? Or are you mainly summarizing what others have said without adding much value? That second question describes most generic blog posts exactly.
Commodity content repeats what already exists. It offers generic advice, adds no perspective, and could carry any competitor’s logo. An AI can replace it instantly because it already knows everything the page says.
Non-commodity content gives the reader something they cannot get from a generic answer. Original analysis. A practitioner’s take. A specific lesson from real work. A framework you built because you needed one.
The test is blunt. Could a competitor publish your exact post with their name on it? If yes, it is commodity content, and AI search treats it that way.
Google even holds a patent built around this idea. Google calls it information gain. It scores how much new information a page adds beyond what is already out there. Google has not confirmed it ranks pages this way. The direction is hard to miss.
Auditing the library: what still earns its place
You cannot fix this post by post in the dark. You need a pass through the whole library that sorts each page into a clear bucket.
Five categories cover most of it:
- Keep: still ranking, still earning clicks, supporting conversions, or holding genuine insight. Leave it alone.
- Refresh: good bones, stale body. Outdated stats, weak examples, no AI-era context. Worth an update.
- Consolidate: three thin posts circling the same topic. Merge them into one strong page and redirect the rest.
- Reposition: generic awareness content that could become specific, opinionated, and buyer-focused with real work.
- Remove or redirect: thin, dated, duplicative, no strategic value. Cut it or point it somewhere useful.
Score each page on a few simple axes: traffic trend, conversion contribution, uniqueness of insight, and overlap with stronger pages. A page low on all four is rarely worth saving.
How to spot content that is truly worthless
Low traffic alone does not condemn a page. Some low-traffic pages still close deals. The real signal is a page with no job to do.
Run a post against this list:
- No meaningful traffic in 6 to 12 months
- No impressions for queries you care about
- No conversions or assisted conversions
- No backlinks
- No original example or proprietary insight
- Mostly generic definitions
- Heavy overlap with a stronger page
- Written for a keyword, never for a reader
A page that checks most of these boxes is dead weight. Keeping it does not help. It dilutes your site and competes with your better work.
What still survives in AI search
The content holding its value shares a pattern. It carries first-hand experience an outside writer could not invent.
That includes expert commentary, industry-specific examples, original data, and strong points of view. Comparison and decision-stage pages earn their keep, because a buyer weighing options needs more than a definition. Real case studies, proprietary frameworks, and detailed buyer guidance do the same.
The throughline is simple. Content that helps a reader make a decision survives. Content that only helps them learn a definition does not.
Rebuilding the strategy for AI search
The old playbook rewarded volume. Publish for every keyword, stack the library, watch traffic climb. That math broke.
The new playbook rewards value. Stop publishing a post just because a keyword exists. Build around the questions that sit close to a decision:
- Comparison and “best option” queries
- Specific client pain points
- Proprietary insight and original research
- Expert-led commentary
- Topics where you hold real authority
Write each piece to serve two readers at once. The human deciding whether to trust you, and the AI system trying to understand and summarize your expertise. Do the first well and the second tends to follow.
Resist the urge to spin up a fresh page for every keyword or AI variation. Google’s scaled content abuse policy targets exactly that move. One strong, comprehensive page beats ten thin ones chasing slightly different phrasings.
What this means for B2B teams
B2B buyers still research. The path just changed. Basic awareness content no longer earns the click it once did, because the assistant handled the awareness step.
The opening sits later in the process. Buyers need help comparing, evaluating, and trusting. So the work shifts to sharper service pages, stronger proof, and clearer points of view. Content that reflects work you have actually done.
A large blog full of generic explainers is not an asset there. It is a maintenance cost.
How Verta helps
Verta works with companies stuck in exactly this spot: a big library, declining returns, and no clear read on what to keep.
We run content and SEO audits, find the commodity pages dragging you down, and surface the gaps worth filling. From there we consolidate weak posts and refresh the ones worth saving. Then we plan non-commodity content built to earn visibility across both traditional search and AI search. The technical side comes with it: structure, crawlability, and internal linking, alongside AEO, GEO, and LLM optimization.
The goal is a smaller, stronger library that works instead of a large one that sits there.
The future belongs to content with a reason to exist
AI search did not kill content. It killed content that adds nothing new.
The brands that win next will not publish the most. They will publish the clearest, most useful, most original work in their category, and let the rest go.
Start with an honest audit. Find the pages doing real work. Be willing to retire the ones that no longer do. Want a second read on what your library is actually worth? Talk to Verta about modern SEO and AI search visibility.
