Google: AI Mode loses to full SERPs on browsy queries

What happened

Per Search Engine Journal, Google's Head of Search Liz Reid has publicly drawn a line between the kinds of queries that belong in AI Mode and the kinds that still belong in a classic ten-blue-links SERP. Her framing: AI Mode wins on complex, multi-step, follow-up reasoning. Full SERPs win on "browsy" queries, the exploratory searches where users want to scan options, compare sources, and click around.

This is the clearest signal yet that Google is not planning to collapse all search behaviour into a single generative interface. Reid's comments, made during a recent podcast appearance covered by Search Engine Journal, explicitly position AI Mode as a complement to traditional search, not a replacement.

The practical translation: Google is building a two-lane highway. One lane is generative, conversational, and answer-shaped. The other lane is link-shaped, scannable, and exploratory. The query itself decides which lane the user lands in.

Why it matters for your brand

For B2B brands, this changes the targeting logic of every content investment. You are no longer optimising for "Google." You are optimising for two distinct retrieval contexts inside Google, each with different winning formats.

Consider a CMO at a multilateral institution publishing research on climate finance. The browsy query ("climate finance reports 2025") still routes to a full SERP, where domain authority, freshness, and clean metadata determine whether your PDF or landing page ranks. The complex query ("how does blended finance compare to concessional lending for adaptation projects in small island states") routes to AI Mode, where the model synthesises across sources and cites a handful. The first battle is won with traditional SEO. The second is won by being the source the model trusts to summarise from. These are not the same skill.

Financial services brands face the sharpest version of this split. A retail-curious user searching "best business savings account" is browsy: they want to compare, click, and self-serve. A treasurer searching "implications of Basel III endgame on regional bank liquidity coverage ratios" is in AI Mode territory. The first audience sees your paid placement and your organic listing. The second audience sees a synthesised answer that may or may not name you. If your thought leadership is gated, unstructured, or buried in PDFs that models cannot parse cleanly, you lose the second audience entirely, and that is the audience with the budget.

For major industrial groups, the split maps onto the buyer journey. Top-of-funnel discovery ("cement decarbonisation suppliers") stays browsy and SERP-shaped. Mid-funnel evaluation ("how does calcined clay cement compare to GGBS for high-rise construction in seismic zones") shifts to AI Mode. Brands that have spent a decade winning the discovery query now need a parallel content layer aimed at the synthesis query. That layer looks different: it is structured, comparative, source-rich, and written for extraction rather than for clicks.

For philanthropic and policy institutions, the implication is about citation share. Foundations and think tanks publish to influence decisions, not to drive transactions. AI Mode is where influence now compounds, because complex policy questions are exactly the queries Reid says AI Mode is built for. If your white papers are not being surfaced when a programme officer asks AI Mode about evidence on cash transfers in fragile states, your influence is invisible at the moment of decision. The browsy SERP still serves your existing audience. AI Mode determines whether you reach the next one.

The content strategy implication is concrete: stop producing one type of asset. Browsy queries reward scannable hub pages, comparison tables, and category landing pages. Complex queries reward deeply structured argumentation with clear claims, named methodologies, and explicit comparisons. The same topic now needs two treatments.

The signal in context

Reid's distinction lands at a moment when most publishers and brands have been treating AI search as an existential threat to organic traffic. The data on AI Overviews shrinking click-through rates is real, and it has driven a defensive posture across enterprise marketing teams. What Reid is describing is more nuanced: Google itself is segmenting its product so that the link economy survives for a meaningful share of queries, while the answer economy takes over for others. For brands, this means the right strategic question is no longer "how do we survive AI search." It is "which of our queries are browsy, which are complex, and are we resourced to win both."

The broader pattern across the major LLM platforms is convergent. OpenAI's ChatGPT search, Perplexity, and Anthropic's Claude with web access all show stronger citation behaviour on complex, reasoning-heavy queries than on shallow exploratory ones. Google saying the quiet part out loud, that exploratory search still belongs to traditional SERPs, gives B2B marketers permission to keep investing in classic SEO for the top of the funnel while building a separate, structured, model-friendly content layer for the consideration and decision stages where AI synthesis now dominates. The brands that split their content operation along that seam in 2025 will own the citation share when the seam widens.