A Single Search Box Is Rewriting the Food Chain

Opening

Dear subscribers, this past Tuesday, the search box changed at Google I/O 2026 — reportedly the biggest overhaul in 25 years. The search bar got longer, you can now attach photos and videos, and AI automatically figures out your intent and generates an answer. There’s even an “information agent” now that roams the web while you sleep, gathering updates for you. I considered doing a full recap of Google I/O 2026, but there are already plenty of people covering that ground. So instead, I want to focus on the one part that struck me most: search.

But watching this announcement, something else occurred to me. Google is the absolute ruler of the search ecosystem. Publishers create content, Google indexes it, and connects it to users — that’s the structure. Hundreds of thousands of media outlets and blogs have survived on top of this food chain. To cut to the conclusion: what Google is doing right now is restructuring that food chain. And the content market isn’t ready to absorb that restructuring yet.

For reference, I’ve criticized GEO/AEO before. In the end, it’s just a downgraded version of SEO, and this announcement pushes that argument closer to confirmed fact. If you’re hoping for organic search traffic, you’ll find more peace of mind focusing on conversion rather than traffic volume going forward. Search-driven traffic is only going to keep shrinking.

🔍 The Biggest Search Box Upgrade in 25 Years — What’s Changing

There are three key points from this announcement.

First, the search box itself has become an AI gateway. The old search box was a tool where you typed keywords and got 10 links back. Now, you type a long sentence, and AI parses your intent — accepting not just text but images, files, videos, and even Chrome tabs as input. Queries in AI Mode reportedly run about 3x longer than traditional search queries. The keyword era is giving way to a conversational one. (It’s sentences now, not just words!)

Second, the line between AI Overviews and AI Mode has disappeared. If you ask a follow-up question on the AI summary (AI Overviews) that appears atop search results, it automatically switches you into conversational search (AI Mode). Users no longer need to be aware of which mode they’re in. Liz Reid, Google’s VP of Search, called this “friction elimination.”

Third, “information agents” have arrived. Set a topic you care about, and AI monitors the web around the clock in the background, alerting you when something changes — stock moves, new real estate listings, concert tickets, that kind of thing. Think of it as an AI-powered version of Google Alerts, except the key difference is that the agent keeps working even after you’ve left Google.

AI Mode surpassed 1 billion monthly active users just one year after launch, and query volume is more than doubling every quarter. Queries in AI Mode run about 3x longer than traditional search. Instead of “best Italian restaurant in Boston,” people now ask things like “a family vacation spot within 4 hours of Philadelphia with beaches and hiking, budget ₩3,000,000.” The very grammar of search is changing.

Google frames these numbers as proof that “people are searching more thanks to AI.” That’s true. But there’s something worth examining here.

🕸️ The Food Chain Is Breaking

The Google search ecosystem isn’t just a tech platform — it’s an economic ecosystem. Publishers (people who post content online) create content → Google indexes it → users search → they click the link → publishers earn ad revenue. This cycle has propped up the web economy for over 20 years.

AI search breaks the key link in this cycle. Users simply have no reason to click through anymore.

Pew Research Center tracked 68,000 real search queries. When AI Overviews appeared, users clicked through to search results only 8% of the time. Without AI Overviews, that figure was 15% — nearly cut in half. And the click rate on source links embedded inside AI Overviews? Just 1%. Displaying a source and actually sending traffic to it are two completely different things.

The Reuters Institute’s analysis of Chartbeat data is even more direct. Over the one year from November 2024 to November 2025, Google search traffic to publishers worldwide dropped 33%. In the US, it fell 38%. Referrals from Google Discover¹ also declined 21%.

The damage isn’t distributed evenly. Small publishers have taken the biggest hit. According to Chartbeat’s breakdown by publisher size, small publishers saw search referrals drop 60% over two years, versus 22% for large publishers — nearly a 3x gap. The less resources an outlet has to build alternative channels, the more concentrated the damage becomes.

There are concrete examples too. The UK’s DMG Media (publisher of the Daily Mail) reported click-through rates falling by as much as 89% on queries where AI Overviews appeared. The travel blog The Planet D lost 90% of its traffic and shut down entirely. Business Insider saw organic search traffic drop 55% between April 2022 and April 2025, and cut 21% of its staff.

Helen Havlak, publisher at The Verge, put it this way: “The extinction-level event has already started, and small publishers have already gone out of business.” Part of the reason is that AI weighs a source’s credibility when deciding what to cite. Whatever outlets we perceive as “well-known,” AI tends to perceive the same way.

The industry outlook is equally grim. In a Reuters Institute survey of 280 media leaders across 51 countries, publishers expect search traffic to fall by an additional 43% over the next three years. One in five respondents expects a decline of more than 75%.

Meanwhile, could traffic from AI platforms fill the gap? Referrals from ChatGPT to publishers are growing fast, but they still account for just 0.02% of total publisher traffic. Even combining every AI platform gets you to only 1%. That’s nowhere near enough to offset the 33% lost from search.

But there’s a structural contradiction here. The content that AI search summarizes is created by these very publishers. If the people making the content disappear, what exactly will AI be summarizing?

💰 Ad Revenue Is Flowing Toward Google

This restructuring of the food chain also involves a restructuring of money flows. Up to now, Google’s search ad business worked like this: send users to publishers and collect ad fees along the way. When a user landed on a publisher’s content, ads via AdSense or Ad Manager generated revenue, and a share of that flowed back to the publisher.

AI search changes that flow. If users get their answer inside Google and leave, they never reach the publisher’s site — the ad impression opportunity vanishes entirely. Instead, Google has started inserting ads directly into AI Overviews and AI Mode. It’s a new ad product that considers both the query and the context of the AI’s answer, enabling more sophisticated targeting than traditional keyword ads.

The numbers bear this out. In Alphabet’s Q1 2026 earnings, search ad revenue on Google’s own properties grew 17% year-over-year to $63 billion. Meanwhile, Google Network revenue — the share distributed to publisher partners — fell 4%. Some analyses note that, for the first time, Google’s own-platform share of total ad revenue has crossed 90%.

To sum up the structure: search queries are hitting all-time highs, and Google’s search ad revenue keeps growing. But the share of that growth flowing back to publishers keeps shrinking. The pie is getting bigger — it’s just being sliced differently.

🇰🇷 What Does This Mean for Korea?

Korea has its own search ecosystem in Naver (South Korea’s dominant homegrown search engine), so Google’s shift might feel like someone else’s problem. But look at a few underlying trends, and that assumption starts to fall apart.

First, Google’s presence in Korea is growing fast. Figures vary by measurement source, but according to StatCounter, Google overtook Naver for the first time in September 2025 (Google at 49.6% vs. Naver at 40.6%). By Internettrend (a Korean web-analytics service), Naver still leads with 62%. The gap comes down to methodology, but what matters is that Google’s share is trending upward.

Second, Naver is moving in the same direction. Naver’s “AI Briefing” feature is structurally identical to Google’s AI Overviews. It shows an AI summary atop search results, reducing the need for users to click through to original content. Creators who monetized through Naver Blog, small merchants who drove traffic through Smart Store[^2] — this entire ecosystem sits within AI search’s blast radius.

Third, business models that depend on search traffic need to change. This isn’t just a publisher problem. Hospitals, private academies (hagwon), online shops — anyone who drove traffic through blog marketing. Tens of thousands of businesses in Korea that run on search traffic fall into this category.

The direction of the response is clear. Shift from a “model that depends on search referrals” to a “model that builds direct relationships.” That means forging direct connections with readers and customers through newsletters, owned apps, communities, YouTube — channels you control. In fact, the Reuters Institute report found that the publishers weathering the search traffic decline best were the ones with a high share of email newsletter and direct-visit traffic.

Search isn’t going away. But the era when search traffic flowed into your business as a given is coming to an end.

Oz’s Lens

Watching all this, I keep coming back to a pattern I’ve seen repeatedly while building go-to-market strategies: the “harvest acceleration” pattern.

Once a platform moves past its growth phase, it starts absorbing directly the value that used to flow to its ecosystem partners. Short-term metrics go up — usage rises, time-on-platform rises. But over the long run, the supply base thins out. As the incentive to create content shrinks, the value the platform itself delivers eventually declines too.

Google’s own reported figure — a 10%+ increase in search usage on queries where AI Overviews appear — is a perfect example. That number proves people use a feature when you show it to them. It proves nothing about whether the answers were accurate. Usage and quality are different metrics, and Google is only disclosing the former.

Traditional search’s “10 blue links” were never elegant. But they served an important function: they let users compare sources, spot conflicting claims, and reach their own judgment. Friction was judgment. As AI search eliminates that friction, it’s also taking away the responsibility of judgment from the user.

Closing

To sum up:

• Google is turning search from an “information-connecting tool” into an “information-delivering tool.” This isn’t a redesign of the search box — it’s a restructuring of value flow across the entire ecosystem.
• In the process, publishers’ traffic and revenue base are shrinking fast. Global search traffic dropped 33% in a single year, and the industry expects a further 43% decline over the next three years.
• AI search accuracy still doesn’t justify this shift. 60% of citations are inaccurate, and 45% of news responses contain significant errors. (Of course, accuracy will keep improving from here.) The era when search traffic arrived as a given is ending. The question to ask now isn’t “how do I rank higher in search,” but “how do I build a structure where customers find me even without search.”

References & Further Reading

Primary sources

• Klaudia Jaźwińska, Aisvarya Chandrasekar, “AI Search Has a Citation Problem,” Columbia Journalism Review (Tow Center for Digital Journalism), March 2025. A landmark study testing citation accuracy across 8 AI search tools with 1,600 test cases.
• EBU & BBC, “News Integrity in AI Assistants,” October 2025. A large-scale study on AI news accuracy involving 22 public broadcasters across 18 countries.
• Reuters Institute/Chartbeat, “Journalism, Media, and Technology Trends and Predictions 2026,” January 2026. Contains the Chartbeat data tracking a 33% drop in search traffic across 2,500 publishers.
• Pew Research Center, AI Overviews CTR study, July 2025. Tracked 68,000 real search queries to measure how click-through rates change with and without AI Overviews.

Background

• Google’s official blog, “Search at I/O 2026,” May 2026. The official announcement of this year’s search updates at Google I/O.
• AdExchanger, “The AI Search Reckoning Is Dismantling Open Web Traffic,” January 2026. A comprehensive look at how AI search is affecting the publisher ecosystem.

The author, Kwangseob Ahn, is a professor of business administration at Sejong University and lead consultant at OBF (Oswarld Boutique Consulting Firm). He teaches statistics and data analysis — business data management and business analytics — while leading GTM and AI strategy consulting in the field, designing the seam between technology and business. He has published academic research on a memory architecture for AI dialogue systems (HEMA) and runs Daily Arxiv, a daily curation of global AI papers. He holds a master’s from Korea University’s Graduate School of Technology Management and a KMBA. He is the author of Homo Brainless: The People Who Outsource Their Thinking.

Footnotes

1. Google Discover: A feed within the Google app or on Android home screens that automatically recommends content based on a user’s interests. Because articles appear without any search action, it’s an important traffic source for publishers. ↩