What Should Publishers Actually Do When AI Is Eating Their Traffic?

The instinct is understandable. Traffic drops, so you add more ad slots. Revenue per session falls, so you push floors higher and load heavier. It feels like a rational response to a shrinking pie.

It is the wrong move — and in most cases, it accelerates the damage.

The traffic contraction publishers are experiencing right now is real and well-documented. Google search referrals to publishers fell roughly 33% globally in the year to November 2025, and individual publishers have reported losses far steeper than that average. Business Insider saw organic search traffic fall 55% over three years. HubSpot estimated a 70–80% drop. But the headline numbers obscure something important: the impact is highly uneven, with publishers most reliant on commodity content — product reviews, lifestyle guides, and how-to articles — taking the hardest hit. These are exactly the content types AI can summarize and serve directly on the results page without ever sending a user to your site.

That distinction matters for how you respond. Not every publisher is facing the same problem, and generic advice — write better content, improve UX, diversify away from Google — treats every site as though it has the same exposure, the same audience, and the same options. Most don't.

Your Site Is Not Every Site

The publishers in the most immediate danger are those whose primary content function is answering quick questions. Informational queries, definitions, how-tos, explainers, study guides, now have a 74% zero-click rate when AI Overviews appear. The user gets the answer. They never visit. The content that drove millions of sessions annually has been, effectively, scraped and replaced by the engine that used to send you traffic.

Chegg is the clearest case study: the education platform reported a 49% decline in non-subscriber traffic as AI Overviews began answering the homework and study questions that had defined its traffic model. DMG Media documented drops as steep as 89% for certain query types.

By contrast, publishers whose content is experiential, community-driven, or requires ongoing access, niche verticals, investigative journalism, tools, interactive resources, are holding up considerably better. A Graphite and Similarweb analysis of 40,000 of America's top websites found organic Google traffic was down just 2.5% overall, with the top 10 US sites actually seeing a 1.6% traffic increase. The losses are concentrated, not universal.

So before you change anything, accurately diagnose what kind of publisher you are. What percentage of your content is genuinely answerable by an AI Overview? What are readers coming back for when they return, rather than landing through search? The answers to those questions should shape every decision that follows.

Know Who Is Actually Coming to Your Site

Most publishers have a rough sense of their total audience. Far fewer have a real understanding of their returning segment — who they are, what keeps bringing them back, and how that group differs from first-time search visitors who may never return.

That distinction has always mattered for monetization. It matters more now. If your search-driven traffic is contracting, the segment that returns by choice is the one with actual long-term value. They're likely more engaged, more receptive to mid-funnel advertising, and more likely to convert on premium offerings. Data from Similarweb suggests visitors arriving through AI search results convert more than twice as often as standard organic visitors — a reminder that volume and value are not the same thing.

Publishers who understand their returning audience can make better decisions across the board: what content to invest in, which ad formats and placements to prioritize, which partners to work with, and how to structure their monetization around actual user behavior rather than aggregate session counts. Publishers who don't have that picture are flying blind at exactly the moment when clarity matters most.

More Ads Is Not a Revenue Strategy

This deserves to be said plainly because the instinct is so common: adding ad density to declining traffic does not offset revenue loss. It makes the existing problem worse.

A reader who found your site through an AI-surfaced link is already approaching with lowered expectations. They came because the AI didn't fully answer their question, or because they wanted to verify something, or because a link was cited and they followed it. If the page they land on is slow to load, visually cluttered, and plastered with intrusive formats — pre-roll mid-article, sticky units obscuring content, multiple auto-refresh slots — they confirm the worst and they don't return.

You've burned the relationship for a fraction of a session's ad revenue.

Publishers who excelled at converting visitors — regardless of traffic volume — demonstrated far greater revenue resilience than those focused purely on traffic acquisition. That pattern holds at the ad product level too. Higher-quality monetization on a smaller but genuinely engaged audience consistently outperforms aggressive monetization on declining volume. The math works because CPMs on engaged users are higher, brand advertiser interest follows audience quality, and you're not actively destroying the retention that's left.

Good UX is not a nice-to-have in this environment. It is part of your monetization strategy, because the users who stay long enough to generate real revenue are the ones who found your site worth visiting in the first place.

Content That AI Cannot Replace

The strategic implication of AI Overviews is worth stating clearly: top-of-funnel informational content is now primarily a citation play, not a traffic play. Teams producing "what is" and "how to" articles primarily to drive blog traffic are measuring the wrong outcome.

That doesn't mean abandoning informational content entirely — it means producing it with the understanding that the primary return is brand visibility and citation in AI answers, not sessions. The content that drives sessions now needs to be something AI cannot adequately replicate: original reporting with sources only you have access to, community or contributor content that reflects lived experience, tools and calculators that require a visit to use, and content with enough recency and specificity that a general AI summary would be insufficient.

Publishers who are still planning their editorial calendars around keyword volume without accounting for AI capture rates are doing the equivalent of optimizing for a channel that has materially changed its terms of service. The planning inputs have changed.

The Monetization Stack Has to Match the Strategy

A publisher can fix their content mix and improve their UX and still have those gains undermined by a monetization stack that's working against them — one that defaults to density over precision, or applies uniform floor strategies across pages with fundamentally different audience value.

This is where the partner relationship becomes the deciding factor. Most monetization partners are optimizing for aggregate yield across their book of business. The tactics that make sense at scale for one publisher may actively harm another. A floor strategy calibrated for a high-volume general interest site will behave differently on a lower-volume niche publication with a premium audience — and not always better.

Aditude's approach is built on the premise that each publisher's setup is distinct and needs to be treated as such. Insights surfaces the data that tells you what's actually happening with your inventory — where yield is leaking, where floors are set wrong, where demand partner configuration is leaving money on the table. Cloud Wrapper gives your ad ops team the infrastructure to act on that information intelligently rather than reactively. And AWP introduces an agentic layer that can monitor, recommend, and execute changes at a speed no manual ops workflow can match.

The philosophy behind all of it is the same: more money should not automatically mean more ads. It means smarter monetization of the audience you've earned, with tooling that can tell the difference.

If your traffic is contracting and you're not sure whether your monetization setup is helping or compounding the problem, that's the right question to start with. Talk to Aditude.