AI answers 68% of local searches now. Most location pages cannot be cited.

Your local traffic has felt off lately.

Fewer calls. Fewer clicks. Customers getting answers and moving on before they ever reach your site.

That is not a slow season.

That is AI Overviews doing exactly what Google built them to do.

What changed

A new Search Engine Journal piece lays out the number.

AI Overviews now appear in 68% of local searches.

Local packs? Only 39% of the same queries.

That is a 29-point gap where your customers are getting answers without ever seeing the map results you have spent years ranking in.

The gap is not even across query types.

Simple transactional searches like “tacos San Francisco” still lean on the local pack.

But informational searches like “how long does an eye exam take near me” trigger AI Overviews 92% of the time.

Hybrid intent searches like “average cost of dental implants in Phoenix” hit 97%.

If your business answers questions, not just fills orders, you are almost entirely living in AI Overview territory now.

Why this matters

The old playbook was predictable.

Optimize your Google Business Profile. Stack citations. Build local keywords. Protect your Map Pack spot.

That system worked for years.

In 2026, it is not enough. Businesses are ranking well in the local pack and staying completely invisible in AI answers for the same queries. Those two things used to be the same game. They are not anymore.

Your competitive set just got bigger.

You are no longer just competing with the business down the street. You are competing with anyone who can answer the customer’s question better than you can.

The piece is also blunt about the implication: AI Overviews control the user experience, not your website. If you only measure traditional local rankings, you have a serious blind spot.

The cookie-cutter location page is dead

Across dozens of accounts, the pattern is the same.

Same structure. Same generic copy. Just swap the city name.

It was never a great strategy. It used to produce okay results. Today, AI systems spot the template immediately and move on.

LLMs are not reading your pages like a human skimming a website. They are looking for fact density. Structured, specific, directly useful information that answers real questions without requiring interpretation.

A 200-word page that says you “serve Nashville with pride” and repeats your service list gives the AI nothing to work with.

What an AI-citable location page actually needs

Three things show up in every page the piece identifies as performing in AI Overviews.

1. Fact density, not keyword coverage

Every location page needs:

• NAP, service area, and hours in a table
• Testimonials from real customers in that specific city, with neighborhood references when possible
• FAQs specific to that market: permit questions, local regulations, climate-specific concerns
• Real project examples or case studies from that location
• A structured data grid showing regional pricing, service tiers, or compliance details

Tables and FAQ schema matter here. Large language models are pattern-recognition systems. Structured formats give AI discrete, labeled data points it can extract with confidence, instead of forcing it to interpret paragraphs.

A page that says which pests show up in Austin in the summer, your average response time, and what the process looks like from first call to completion is the page that gets pulled into the Overview.

2. Geographic legitimacy through landmark images

Google’s vision AI reads your photos to verify you actually belong to the place you claim.

Generic stock images of smiling professionals in a generic office read as low-context boilerplate. The AI cannot tell where you are. When it cannot tell, it cites someone else.

The fix is intentional local imagery. Three landmark images per location page, sourced through Google Images Creative Commons filters:

• Search the specific landmark by name (“Navy Pier,” not “Chicago”)
• Use Tools to filter for Creative Commons licenses
• Click through to verify the license type
• Rename the file with a descriptive name like “navy-pier-chicago-lakefront.jpg”
• Write alt text that describes the image and mentions the location naturally
• Attribute in a small caption when the license requires it

When your alt text says “Navy Pier at dusk,” your filename is descriptive, your copy references Chicago residents, and your quick-facts section lists Illinois-specific regulations, you have built a page AI can locate.

3. Entity consistency across the public web

This is the part the piece does not name explicitly, but every cited page already has it.

If your website, Google Business Profile, directories, and schema disagree on your service area, your hours, or your specialties, the AI has to guess. When it guesses, the cleaner competitor wins.

For most local businesses, the entity-clarity work is the highest-leverage move before any new content gets written.

What SMBs should do now

Start with three checks, in this order.

1. Pick the three location pages that matter most to revenue. Do not try to rebuild fifty pages at once.

2. Audit one against the fact-density list above. Count the specific, locally-answerable facts on the page. If the count is under ten, you know the issue.

3. Replace the generic photos with three real local landmarks. Use the Creative Commons workflow in the piece. Rename the files. Write real alt text.

Then run the same exercise against your Google Business Profile. Make sure services, hours, and service area match the page word for word.

For dental, home services, automotive, and healthcare, this is the work that decides whether AI cites you or your competitor when a customer asks a local-intent question this quarter.

If you want the broader picture on why your Google Business Profile is now the front door for most of those searches, start here.

The takeaway is simple.

Local search is no longer a ranking game.

It is a citation game. And the page that wins is the one AI can actually understand.