AI Overviews Killed Your Top-of-Funnel Traffic. Here's What to Do.

What changed: AI Overviews vs old featured snippets

Featured snippets (the answer box at the top of search results) have been around since 2014. They cost some clicks but felt manageable: people still scrolled, still clicked, still arrived on the publisher's page. AI Overviews are different in three meaningful ways:

1. The answer is fuller. Featured snippets pulled one paragraph; AI Overviews synthesize 3–6 paragraphs from multiple sources, often answering the question completely.
2. The user often doesn't scroll past it. Eye-tracking studies in 2025 showed 40–60% of users with informational queries don't move past the AI Overview. The traditional 10 blue links sit below the fold for most queries.
3. Citations are aggregated. An AI Overview that cites you alongside five other sources gets 1/6 of the click attention — not 100% as a featured snippet would have.

The composite effect: top-of-funnel queries that used to drive 100,000 clicks/month to a B2B publisher now drive 30,000–50,000. The traffic didn't go to a competitor — it stayed on Google.

The 5 B2B traffic patterns hit hardest

Not all traffic is affected equally. The patterns we see across B2B clients:

1. “What is X?” queries (definitional content)

The classic top-of-funnel: “What is RevOps?”, “What is product-led growth?”. Down 50–70% in click volume. AI Overviews answer these completely; users rarely scroll.

2. Listicles (“Best X for Y”)

“Best CRM for B2B SaaS”, “Top 10 marketing automation tools.” Down 40–55%. AI Overviews compress the list to 3–5 entries with names, descriptions, and links to vendor sites — bypassing the listicle author.

3. How-to content for general processes

“How to write a B2B email,” “How to build a sales pipeline.” Down 35–50%. Generic how-to is the easiest content for AI Overviews to summarize.

4. FAQ-style queries

“Why is X important?”, “Should I use Y or Z?”. Down 40–60%. Direct-answer queries are AI Overview's home turf.

5. Glossary content

Single-term definitions that historically drove SEO traffic at scale. Down 60–75%. Practically gone for top-of-funnel traffic.

What's not hit hardest: comparison content (X vs Y), tool-specific deep dives, original research with proprietary data, case studies, and bottom-of-funnel intent (“buy X,” “X pricing,” “X demo”). AI Overviews are still appearing on those queries; users still click through because they want the depth.

The new B2B TOFU strategy that actually works

Stop measuring TOFU by clicks. Start measuring it by AI Overview citation frequency, brand-mention frequency in LLM responses, and downstream branded-search lift. This is the shift.

What that looks like in practice:

1. Stop writing 1,500-word definitional posts. The cost-of-writing-vs-clicks-received math doesn't work anymore. Use that budget for the formats below.
2. Write content explicitly designed to be cited. Direct-answer leads, structured data, original research with quotable statistics. More on what makes content citation-worthy.
3. Invest in schema markup. FAQPage, HowTo, Article, and Product schema increase the odds of AI Overview citation. The 8 schema types that matter for B2B AEO.
4. Get cited on third-party authoritative sites. AI Overviews lean heavily on third-party citations and reviews. A mention of your brand on a respected industry publication is worth more than another blog post on your own site.
5. Build mid-funnel content with original data. “The 2026 B2B SaaS RevOps benchmarks” with proprietary data attracts citations from the AI Overviews and the journalists feeding the AI Overviews.
6. Lean into bottom-funnel content. Comparison pages, integration pages, alternative-to pages, pricing pages. Buyers searching these still click through, and the intent is closer to revenue.

What to measure now

The traditional TOFU metrics (organic sessions, time on page, scroll depth) are losing their grip on the question that matters: are the people we want to reach learning who we are? The replacement metric stack:

Most B2B teams we work with are surprised at how stable their pipeline is once they zoom out from raw traffic. Smaller funnel, higher intent, similar revenue — just a different shape than 2023.

What about brands that are losing pipeline, not just traffic?

If your pipeline is dropping (not just traffic), the diagnosis is different. The likely causes:

1. You weren't getting leads from TOFU traffic in the first place. The traffic was generating brand awareness; the leads came from middle / bottom funnel. Audit which content actually produced pipeline; you may discover TOFU was less load-bearing than you thought.
2. Your competitors are getting cited; you aren't. AI Overviews don't list everyone — they pick 3–6 sources. If yours isn't among them, you're invisible. Audit which sources are getting cited on your top queries and reverse-engineer why.
3. You're not appearing in LLM responses at all. Test it directly: ask Claude / ChatGPT / Perplexity the questions your buyers ask. If you're never mentioned, that's the gap.

The fix in all three cases is structural, not tactical: become the source AI surfaces cite, not just another publisher chasing clicks.

What NOT to do

Five reactive moves we've seen B2B teams make that don't help:

1. Don't double down on definitional / glossary content. Writing more 1,500-word “What is X” posts in 2026 is digging the hole deeper.
2. Don't try to game AI Overviews with prompt-injection or hidden text. It doesn't work, and Google's spam team is actively penalizing it.
3. Don't shut down the blog entirely. The blog is your AEO surface area. The format and metrics shift; the channel doesn't go away.
4. Don't replatform your CMS to chase “AEO best practices.” Your CMS isn't the constraint; the content shape is.
5. Don't lay off the SEO team and replace them with “AEO experts” on LinkedIn. AEO is an evolution of SEO, not a replacement of it.

Frequently asked questions