When Google introduced AI-generated summaries into its search results in 2024, it handed the media industry a new vocabulary for an old anxiety. Publishers began calling it the "zero-click era" — a moment in which readers find what they need without ever touching a link. For ad-dependent outlets whose business model runs on page views, the consequences have been severe: traffic has collapsed, and the revenue that once followed it has not found a new home.
Huang Chin-yung, founder of DIGITIMES — a publication dedicated to Taiwan's semiconductor supply chain and the broader Asian technology industry — has watched this unfold from an unusual vantage point. His site's traffic has gone up.
Speaking on a Taiwan FactCheck Center YouTube talk show, Huang offered a diagnosis of why DIGITIMES has moved in the opposite direction from most of its peers, and what the experience implies for publishers still trying to make sense of where journalism goes next.
When Google Broke the Traffic Model
The logic of the zero-click era is straightforward: if a search result now summarizes the answer, a reader has diminishing reason to click through to the source. For outlets producing news that any AI can summarize — commodity wire stories, press release rewrites, aggregated takes on widely covered events — this presents an existential problem. The content is abundant, interchangeable, and easily consumed in a pull quote.
DIGITIMES produces a different kind of content. Its reporting covers original equipment manufacturers, component pricing shifts, chipmaker capacity allocations, and the cascading effects of decisions made by a handful of Taiwanese firms that underpin the global electronics industry. This is information that does not exist in a searchable summary because, for the most part, it does not exist anywhere else.
Why Taiwan's Industrial Dominance Created a Captive Readership
Taiwan's exports are expected to exceed $1 trillion this year, with electronics accounting for more than 80% of that figure. That volume of commerce sustains a professional class — engineers, procurement officers, equity analysts, supply chain managers — whose jobs depend on accurate, timely, granular intelligence about what is happening at the factory level. General-interest technology coverage does not serve them. An AI summary of a Reuters item does not serve them either.
Huang traces the English edition of DIGITIMES to a moment of discovery. He found that professionals in Silicon Valley — people working at the heart of the technology industry — often had no idea where the motherboards inside their own flagship products were manufactured, or which Asian firm had designed a critical chip they were deploying at scale. The knowledge gap was both vast and commercially significant. Nobody was bridging it in English, at a professional standard, at any price.
"Position Asian information correctly, then sell it to the Americans," Huang said. "The English edition was always going to have a market." Today, traffic to DIGITIMES' English and Chinese editions runs at comparable levels — a signal that the hypothesis has held.
A $550 Subscription That Justifies a Premium
Most international media outlets with subscription businesses charge between $100 and $400 per year. DIGITIMES charges $550 — roughly NT$17,000. The gap is not arbitrary. When a semiconductor analyst needs to understand what is happening three tiers deep in a packaging supply chain, the answer does not live in a publicly indexed article. It lives in proprietary sourcing, primary interviews with manufacturers, and reporters who have spent years building relationships inside facilities that rarely speak to the press.
"The internet should not be a place for breaking news. It should be a database." — Huang Chin-yung, Founder, DIGITIMES
That reframing — the internet as archive rather than newsfeed — is central to how Huang thinks about DIGITIMES' durability. If the value of a publication lies in what it has reported, organized, and made searchable over time, then an AI summary of today's headlines is not a competitive threat. It is irrelevant to the question a subscriber is actually trying to answer. DIGITIMES' English and Chinese editions have already reached near-equivalent traffic levels, suggesting the model translates across languages.
Breaking the Traffic-First Mindset
Huang's argument is not simply a defense of his own business model. It is a critique of how the media industry as a whole has framed its purpose. The assumption that large audiences and high traffic are the prerequisite for commercial viability has, he argues, led publishers to produce content designed to be skimmed rather than used.
In its place, he proposes something closer to a service model: find the specific group of people who need something they cannot easily get anywhere else, deliver it with precision, and build an interaction deep enough that the reader considers it part of their professional infrastructure. Taiwan's educated population and near-universal smartphone penetration mean the audience infrastructure already exists. The gap is on the supply side.
Huang illustrated the principle with a deliberately unglamorous example: a market vendor and her regular customers, sharing real-time updates on what is in stock and at what price. The content is hyperlocal, functional, and of no interest to anyone outside that community — yet genuinely useful to the people who need it. That distinction, he argues, is the whole point.
The End of Mass Media and the Question That Follows
The era of mass media, in Huang's view, has already ended in any meaningful sense. Individual content creators routinely reach audiences that rival or exceed those of legacy broadcasters. The audiences that remain with traditional outlets are fragmenting along interest lines that algorithms are better at mapping than editorial calendars. Publishers still chasing the broad middle are running toward a market that is already gone.
What survives, he suggests, is specificity matched to demand. Industry stories that most editors would bury — too technical, too niche, too far removed from the general reader — hold substantial value in a searchable archive for the professionals who need them. The stories that seem commercially unviable in a traffic-driven model are frequently the ones with the highest value in a knowledge-services model.
The question Huang poses to Taiwan's media ecosystem is pointed: the country has produced some of the world's most technically accomplished professionals across engineering, manufacturing, and chip design. Who is building the knowledge infrastructure to serve them? In the zero-click era, the answer — and the business — belongs to whoever can deliver something an AI summary cannot replace.
Social Media Links & Editor Tag - Penny