If your local search traffic has felt off lately, it’s not a technical bug and it’s not seasonal. AI Overviews are answering the questions your customers used to type into Google, and a growing share of them never reach the map pack you’ve spent years optimizing for.
The old local SEO playbook — Google Business Profile, citations, localized keywords, Map Pack rank — still works, but it no longer covers the whole picture. Businesses now routinely rank well in the traditional local pack while being invisible in the AI-generated answer for that exact same query. Those two used to be the same fight. They aren’t anymore.
The Visibility Gap Most Businesses Aren’t Tracking
Independent research from Whitespark puts AI Overviews on 68% of local searches, against just 39% for traditional local packs — a roughly 29-point gap where customers get an answer without ever seeing the map results a business worked to rank in. Some AI Overviews embed a mini local pack; others synthesize a full answer with links and skip the map entirely. Either way, Google now controls that layout, not the business owner, which means map-pack rank alone is no longer a reliable visibility measure.
The gap isn’t uniform across query types. Simple transactional searches (“tacos near me”-style queries) still lean heavily on the local pack. Informational queries — things like how long a service takes — trigger AI Overviews around 92% of the time per Whitespark’s data, and hybrid intent-and-cost queries hit roughly 97%. Put simply: if your business answers questions rather than just fulfilling instant transactions, you’re now operating almost entirely inside AI Overview territory.
Vertical matters too. Legal queries trigger AI Overviews almost everywhere regardless of geography; home services are more mixed; restaurants and retail still lean on the traditional pack for simple searches. The bigger shift is competitive scope: a local business is no longer just competing with the shop down the street. It’s competing with anyone, anywhere, who can answer the customer’s question more clearly.
What Makes a Location Page Worth Citing
The cookie-cutter location page — same template, same generic copy, city name swapped — was a mediocre strategy even before AI search. Now it’s actively invisible. AI systems don’t skim a page the way a human does; they look for fact density: structured, specific, directly usable information that doesn’t require interpretation. A page that says a business “proudly serves” a city and repeats a service list gives an AI model nothing concrete to extract.
Three things build a location page AI will actually pull from: structural fact density, verifiable geographic specificity, and consistent entity data across the web. Here’s what each looks like in practice.
1. Structure for Extraction, Not Just Keywords
Every location page needs, at minimum: NAP, service area, and hours in table form; testimonials from customers in that specific city, ideally with neighborhood references; FAQs specific to that market (local regulations, permit questions, climate-specific concerns); real project examples from that location; and a structured data section covering pricing ranges or service tiers for that market.
Tables and FAQ schema aren’t decorative here — they’re the format AI models can extract with confidence, versus paragraphs it has to interpret. A page listing exact response times, common local issues, and a step-by-step process is a page that gets pulled into an AI answer. Publishing more content without improving that structure works against visibility in an AI retrieval environment; precision matters more than volume.
2. Prove the Location Is Real
Google’s vision systems can assess whether a page’s imagery actually supports its location claim. Generic stock photos of smiling professionals in an unplaced office read as boilerplate — there’s no location signal for the AI to verify, so it looks elsewhere for a source it can trust.
The fix is deliberate local imagery: a recognizable local landmark as the primary hero image, a second image showing people in the area with a landmark visible in the background, and a third full-width shot of a local landmark or skyline. Source landmark photos through Creative Commons search filters on Google Images, verify the license on the original source page before using anything, name the file descriptively (a landmark-and-city filename, not a camera default), and write alt text that names the location naturally — with proper attribution where the license requires it. Three deliberate location signals at the image level, before an AI model even reads the copy.
3. Keep Entity Data Identical Everywhere
AI systems cross-reference a location page against everything else they can find about a business. Inconsistent NAP across directories, outdated hours on one map platform, mismatched service categories between a Google Business Profile and the website — each of those lowers the AI’s confidence in the entity, and lower confidence means a lower citation chance.
Business name, address, phone number, service categories, and description need to match, word for word, across Google, Bing, Apple Maps, and every relevant directory. A quarterly audit is a small time investment that pays off directly in citation consistency across AI platforms.
Don’t Skip the Pricing Page
Cost and pricing queries trigger AI Overviews more than 80% of the time, and most local businesses avoid answering them out of fear of tipping off competitors or locking themselves into a number. That avoidance has a cost: if a business won’t answer “how much does this cost,” the AI finds a competitor who will — and that competitor gets the citation, the click, and the call.
The fix isn’t publishing a fixed rate card. A realistic range, paired with the variables that move the price — size, materials, urgency, local permit requirements — gives an AI model exactly what it needs: a factual, organized resource it can cite confidently for cost-intent searches. Write it in the phrasing real customers use (tools like AnswerThePublic or Google’s “People also ask” are useful here), not the phrasing marketers default to.
Citation Authority Doesn’t Stop at Your Website
A well-built location page is the foundation, not the whole strategy. Research from Omniscient Digital analyzing tens of thousands of branded-query citations found owned content accounts for roughly a quarter of citations — the rest comes from off-site sources. A perfect location page on its own isn’t enough; it needs a wider content environment feeding AI systems context and credibility. Four places to focus:
- Above-the-fold answers on commercial pages — FAQ blocks, cost breakdowns, and short how-tos placed directly on service pages rather than buried in a blog, since a large share of AI citations pull from early-page content.
- Directory alignment — conflicting NAP data across directories erodes AI confidence in the entity; matching data across all of them builds it.
- User-generated platforms — forums and video platforms are increasingly cited sources in AI Overviews; genuine reviews and well-optimized video descriptions with local keywords carry real weight.
- Original local data — a business’s own cost benchmarks, satisfaction data, or market reports can’t be found anywhere else, which forces an AI model to cite the source directly rather than paraphrase a competitor.
Content freshness matters too: research from Ahrefs analyzing millions of AI citations found AI-cited content skews meaningfully newer than traditionally ranked content. Treat location and pricing pages as living documents with a quarterly refresh, not a set-and-forget asset.
The Practical Takeaway
AI-driven local search is still early — these systems are actively building their trusted-source indexes, which leaves a real window to establish citation authority before every competitor catches on. Traditional SEO earns a rank. This work earns a citation. Increasingly, the citation is what determines whether a customer calls or never finds out the business exists at all.