Top-of-funnel content is not dead. But publishing another generic “what is” article and waiting for Google traffic is getting harder to justify.

Search Engine Land analyzed traffic trends across 30 major publishers in nine industries. The results show a split. Some sectors are still getting value from informational content. Others are watching traffic slide as Google AI Overviews and large language models answer more questions directly.

That makes the old TOFU formula much less predictable.

AI Search Is Taking More Informational Clicks

The pressure on top-of-funnel content is easy to see. Search Engine Land notes that organic search clicks have fallen 42% since Google expanded AI Overviews. LLMs are also answering more informational questions inside their own interfaces. That leaves fewer reasons for users to visit a publisher’s website. At the same time, AI tools have made it cheap to produce thousands of informational articles. More pages are competing for a smaller pool of clicks.

TOFU Traffic Is Not Falling Everywhere

The traffic decline is not uniform. Search Engine Land compared August 2026 estimates with August 2025 across 30 large publishers. Finance, healthcare, legal, and consumer technology publishers all saw declines in top-of-funnel traffic. Cybersecurity and marketing software moved in the opposite direction. Cloud infrastructure also finished the period with positive results. The industry itself may now matter more than the funnel stage.

YMYL Publishers Are Feeling More Pressure

Your Money or Your Life topics appear particularly exposed. These include finance, health, and legal information. Google often provides AI-generated answers for these searches because users usually want quick and authoritative information. That can remove the need to click another page. Search Engine Land found declines across major publishers in those categories. Even highly established websites were not immune. Smaller publishers may face an even tougher environment.

B2B Technology Still Has Room for TOFU Content

B2B technology looks different. Buyers researching cybersecurity software or cloud migrations rarely make decisions from one short answer. They often need deeper explanations, technical comparisons, and expert guidance. That gives publishers more room to provide useful information. Search Engine Land found improving TOFU performance across cybersecurity and marketing software publishers. Fast-changing technical subjects may also give LLMs a reason to search the live web and cite newer sources.

Moving Everything to Middle-of-Funnel May Not Fix the Problem

Some marketers might respond by cutting informational content and moving resources toward comparison pages. The data suggests that may not solve much. Search Engine Land compared TOFU pages with middle-of-funnel content built around terms such as “best” and “vs.” Both types of pages generally moved in the same direction within each industry. If an industry is losing organic visibility, changing funnel stages alone may not reverse the trend.

Local Content Still Has an Opening

Local search remains one of the stronger areas for publishers. AI systems still struggle with reliable location awareness. Google also shows AI Overviews less often for many local queries. That keeps traditional search results useful. The opportunity is not just building hundreds of basic city pages. Search Engine Land argues that local content should address specific needs inside each service area. Useful local detail gives users a reason to click instead of accepting a generic AI answer.

Interactive Tools May Be More Valuable Than Another Guide

Calculators, screeners, and other digital tools could become stronger TOFU assets. Generic how-to articles are easy for AI systems to summarize. Interactive tools are harder to replace. Search Engine Land found that plain educational content received LLM citations only around 12% of the time in an earlier analysis. Tool pages performed much better. AI assistants could recommend them by name because users still needed to visit the page to use the function.

Real Expertise Has Become Harder to Fake

Human expertise may also become more valuable as generic content becomes cheaper. AI summaries tend to produce safe and predictable explanations. They struggle to reproduce firsthand experience, strong opinions, mistakes, or nuanced trade-offs. A subject matter expert can bring those elements into an article. That makes the page different from hundreds of summaries covering the same topic. Original insights can also improve the chance of earning citations from AI search systems.

Mass AI Content Looks Like a Riskier Bet

The easiest type of TOFU content to automate is also becoming the hardest to defend. AI can quickly produce definitions, basic guides, and introductory explainers. Publishers can scale that output almost without limit. The problem is that everyone else can do the same thing. Search Engine Land argues that fully automated AI content may offer little long-term value. Google has also continued to target scaled content abuse through spam updates. Using AI in the writing process is not necessarily the problem. Publishing large volumes of interchangeable content is.

AI Detection Is Becoming Part of the Web

Readers and platforms are also getting better at recognizing repetitive AI writing. Search Engine Land points to wider adoption of AI detection and watermarking technologies. Exposure itself matters too. People are becoming familiar with predictable phrasing, structure, and tone. That makes bland AI-generated copy easier to notice. Publishers relying on large volumes of nearly identical articles could struggle to build trust, even if the pages technically rank.

TOFU Content Now Has to Earn Its Place

Top-of-funnel content can still work in 2027. It simply cannot rely on the old formula.

B2B technology, local search, interactive tools, and expert-led articles still show promise. YMYL publishers face a tougher environment. Generic informational pages are also being squeezed from both directions. Competitors are producing more of them, while AI systems are answering many of the same questions without sending users elsewhere.

That changes the standard for publishing.

A page cannot exist only because a keyword has search volume. It needs something users cannot get from a five-line AI summary. Search Engine Land puts the decision in simple terms: marketers should ask whether the page genuinely deserves to be published.

Sources