Your map pin is safe. Your cost page isn't.

Google’s AI Overviews — the machine-written summary block at the top of the results page — had two billion monthly users by the middle of 2025, on Google’s own earnings-call numbers, and a separate chat-style AI Mode opened to every U.S. user that May. How often the block actually appears has been a moving target: Semrush tracked it triggering on 6.5% of queries in January 2025, peaking near 25% mid-year, and settling around 16% by November — while steadily expanding out of trivia and into commercial and transactional searches. The shape of the query matters most: Pew found 60% of question-word searches produced an AI summary, against 8% of one- and two-word searches. Long, spoken, question-shaped searches are the AI’s trigger — and that is exactly how families ask about death.

The split that decides a funeral home’s exposure

For local businesses, the honest picture isn’t “AI everywhere.” It’s a clean split by intent. Whitespark’s local study found AI Overviews on just 15% of transactional local queries — the “near me” kind — where the classic map pack still showed 93% of the time. Flip to informational queries and the numbers invert: an AI Overview on 92%, the local pack on 6%. Hybrid questions (“cost of X in [city]”) triggered the block 97% of the time.

No study has isolated death-care specifically — worth saying plainly, since anyone quoting a measured “funeral AIO rate” is inventing one. But map the pattern onto a funeral home’s actual query book and the exposure sorts itself. “Funeral home near me” and “[town] funeral home” are transactional: still a map-pack game, the one your Google Business Profile already plays. “How much does cremation cost,” “what is direct cremation,” “how do prepaid funerals work” — the questions a family asks first, days before they search for a building — are informational, and informational is the AI’s home turf. Your storefront is defended. Your education and price content is where the exposure lives.

What the answer block costs — and the fight over it

When the block appears, the clicks change. Pew’s browsing-panel data: users clicked a traditional result on 8% of searches with an AI summary versus 15% without, clicked a source link inside the summary just 1% of the time, and simply ended their session 26% of the time versus 16% without. Ahrefs measured the same force from the ranking side: the number-one result lost about a third of its click-through under an AI Overview in early 2025, and by its December data the loss had widened to 58%.

Google disputes the framing, on the record — calling the Pew study “a flawed methodology and skewed queryset” and saying it has “not observed significant drops in aggregate web traffic” — and claims the Overviews drive 10% more searches on the query types that show them. Nobody outside Google can audit that; the independent measurements and the platform’s assurance sit unreconciled. What an owner can bank either way: for the question-shaped queries, more families now read an answer assembled by Google than read any single page, including the #1 page.

Being cited is a different game than ranking

Here’s the part that rewrites the old playbook. Ahrefs’ March 2026 analysis of four million AI Overview citations found only 37% of cited pages also rank in that query’s top ten organic results — and more than a third come from pages that don’t rank in the top 100 at all, pulled in through the “fan-out” searches Google’s AI runs behind the scenes. (Ahrefs’ mid-2025 figure was ~76%, but that run counted only each Overview’s three most visible citations — the drop is partly measurement, not only Google.) Ranking #1 no longer guarantees the citation, and not ranking no longer excludes you. The block quotes whoever answers the question most quotably.

That cuts both ways for the industry. The broadest measurement yet — SOCi’s 2026 Local Visibility Index, 2,751 multi-location brands across more than 350,000 locations — found ChatGPT, its most selective platform, recommending brand locations just 1.2% of the time (Gemini: 11%), against a 35.9% average appearance rate in Google’s classic map pack. Those are chains with listings teams — and SOCi’s own reading is that the scale hurts them: AI systems, it notes, often favor single-location businesses, whose signals are simpler and more unified. Read as direction rather than gospel, it says the answer block today belongs to almost nobody — it goes to whoever bothers to be quotable. (The measured numbers, with their limits, are collected in one page we keep current.) And the counterweight is the one lever peer-reviewed research supports: pages with cited statistics and sources gain up to 40% in AI-answer visibility, with the biggest lift going to smaller sites. The quotable page is a game an independent can win. (What nobody should sell you: “schema markup boosts AI citations 2.5×” — a vendor number with no traceable study behind it. Structured data helps machines parse the page; the magic multiplier is invented.)

So the strategy is the split, run deliberately. Defend the storefront query with the profile and the map pack — that game is still 93% of the local moment. Win the question queries with pages built to be quoted: the cremation-cost page with real figures, the plain-language planning answers, each one sourced the way the machines reward. The AI Overview didn’t take your customers — it took the introduction. Families still pick a funeral home; they just arrive having already had the first conversation with Google. Build the pages that conversation quotes, and you’re back in the room.