What New SEO Tactics Are Needed for AI-Driven Local Searches?

What Changed in Local Search Results?

The local pack, the three-listing map block that has anchored local search for over a decade, no longer owns the top of the page. Whitespark's study of 540 city-plus-industry query combinations across Houston, Phoenix and Denver, published May 12, 2025, found AI Overviews appearing in 68% of local searches on average, against 39% for the local pack. Search Engine Journal's July 22, 2026 coverage of the same research confirmed the gap has held: a 29-point split between the feature businesses spent years optimizing for and the one now answering the question instead.

The change is not uniform. It runs on query intent, and the split is sharp. Classic 'near me' local-intent searches, the kind tied to a specific transaction like 'plumber near me,' still trigger the local pack 93% of the time and an AI Overview only 15% of the time. Flip to informational queries, like 'how long does an eye exam take,' and AI Overviews trigger 92% of the time while the local pack shows up in just 6%. Hybrid queries that mix intent, like 'average cost of dental implants in Phoenix,' trigger an AI Overview 97% of the time. See the full breakdown below.

The second half of the change is about sourcing, not just placement. Per the same July 22, 2026 reporting, only 23% of AI Overview citations for local queries come from a business's own website. The remaining 77% come from external sources, with Reddit the single most-cited source at 21% and YouTube close behind at 18.8%. Cited content also runs 25.7% fresher on average than content that ranks traditionally, and 44.2% of citations pull from the first 30% of a page. Being the business does not guarantee being the source.

Who Does This Local SEO Shift Affect?

This hits local, multi-location, and service businesses hardest: plumbers, dentists, personal injury lawyers, optometrists, medical clinics, real estate agents, and any category where a customer researches before they call. Whitespark's six-industry sample maps almost exactly onto this group.

It does not hit every query type equally, which is the part worth sitting with. If someone searches 'emergency plumber near me' at 2am, they still see a local pack and a map, and Google Business Profile optimization still earns that visibility. But if the same person searched 'how much does a plumber cost per hour' an hour earlier while deciding whether to call anyone at all, an AI Overview answered that question, and it cited whatever source Google's model trusted enough to quote. A business absent from that citation is invisible at the exact moment a prospect is forming a shortlist.

That is the practical meaning of citation share for a local business: not where you rank, but what percentage of the relevant AI answers in your category and geography actually name you. A business can rank well and still lose the moment that matters if it never gets cited on the pricing, duration, and comparison questions that increasingly route through an AI Overview instead of a results page.

How Should You Respond?

The response splits into two tracks: what you publish, and what gets said about you elsewhere. Both matter, because 77% of citations are external.

On your own site, the priority is structure over prose. One consultant quoted in the July 22, 2026 reporting put it plainly: large language models are pattern-recognition systems, and structured formats give them discrete, labeled data points to extract with confidence, rather than requiring them to interpret meaning from paragraphs.

1. Build location pages with fact-dense, structured content: pricing tables, service-area lists, hours, FAQs, and schema markup, not a single paragraph of marketing copy.
2. Answer the pricing and duration questions directly and specifically, since these are the query types now dominated by AI Overviews (92% to 97% trigger rates).
3. Audit NAP (name, address, phone) consistency across directories on a quarterly cycle, since AI Overviews cross-check entity consistency before citing a source.
4. Use locally specific, properly attributed imagery. Geographic authenticity is part of how these systems judge whether a source is genuinely local.
5. Keep content fresh on a real cycle. Cited content runs 25.7% fresher on average than content that ranks conventionally, so a location page written once in 2023 and left alone is working against you.
6. Earn mentions outside your own domain. Since Reddit and YouTube alone account for close to 40% of citations, a presence in relevant local discussion threads and video content is not optional anymore, it is a citation channel.

What Should You Track Now That the Local Pack Isn't the Whole Story?

Ranking position in the local pack no longer tells you whether you are visible where it counts. Track presence separately by query type: 'near me' transactional searches where the pack still dominates, and pricing or comparison searches where an AI Overview is answering 9 times out of 10.

For the second category, the useful measurement is not rank but citation share, the share of relevant AI answers that name your business, tracked across ChatGPT, Perplexity, Gemini, and Google AI Overviews, not just Google's classic result page. A business that only watches map-pack position is measuring the 39% of queries where the old rules still apply, and missing the 68% where a different game is already being played.

Sources

AI Overviews appear in 68% of local searches on average versus 39% for the local pack, with a strong inverse relationship by query intent (local-intent: 15% AI Overview vs 93% local pack; informational: 92% vs 6%; hybrid: 97% vs 17%), based on 540 queries across Houston, Phoenix and Denver in six service industries Whitespark, 2025-05-12

77% of AI Overview citations for local queries come from external sources rather than the business's own site, with Reddit the top-cited source at 21% and YouTube at 18.8%; cited content runs 25.7% fresher on average and 44.2% of citations pull from the first 30% of a page Search Engine Journal, 2026-07-22