The Missing Page Problem: Why AI Sends Buyers to the Wrong Store

What we observed

A multi-location brand we'll call The Brand runs two stores under one corporate identity. Two markets, two cities, ~80 miles apart. Same parent company, separate websites, separate customer bases. We'll call them Location A and Location B.

The Brand asked us to test how the major AI assistants were recommending Location B in Location B's own market.

We tested ten high-intent shopper queries — the kinds of questions a customer types into ChatGPT or Gemini when they're ready to spend money. *Best place to buy. Service near me. Lease deals. Trade-in.* That kind of thing.

Same brand. Same parent company. Different store. Wrong answer for the shopper.

What was actually happening

When you operate two stores under one brand, traditional SEO logic says they should both rank fine in their own markets — Google's local algorithm handles it. In Google Search, that's largely true.

Generative AI doesn't work like traditional search.

When Gemini or Perplexity answers a shopper's question, it doesn't ask *"which dealer ranks highest in that city."* It scans the entire web presence of the brand, finds the single best-matching page for the specific query, and cites that page. Whichever store's website hosts that best-matching page wins the answer.

Here's the simplest way to picture it. Imagine a Honda dealer group that operates one rooftop in Austin and one rooftop in San Antonio (we're using neutral examples so it's easy to follow). The Austin store's website carries pages with titles and URLs like:

"Honda Civic Lease Prices Near San Antonio" > "Honda Dealer Near San Antonio"

Those pages live on the Austin store's website, but every signal on them — the page title, the URL slug, the on-page copy — is targeted at the San Antonio market. The Austin store built them years ago to capture spillover searches from the neighboring metro.

The San Antonio store's own website? Nothing equivalent. Its homepage said *"serving the San Antonio area"* and that was it. No page targeting any specific neighborhood. No page naming any city explicitly.

So when a buyer in San Antonio asked Gemini *"what's the best Honda dealer near me?"*, the AI scanned the brand's web presence and found exactly one page that answered the question: the Austin store's *"Honda Dealer Near San Antonio"* page. It cited that page. The shopper was sent to a store 80 miles away — inside their own brand.

Not because the AI made a mistake. Because the San Antonio store had nothing for the AI to cite.

That's the Missing Page Problem.

Why this matters for the business

Three things every operator should understand about why this problem is structural, not cosmetic.

In the case above, three out of ten high-intent queries in Location B's home market were sending shoppers to a sister store 80 miles away. That's not a hypothetical — those are real shoppers, asking real buying questions, getting routed to the wrong place. Some convert at the wrong store. Most just go elsewhere when the answer doesn't match what they were looking for.

Three questions every multi-location operator should ask

If your business operates more than one location under shared branding, sit with these three questions. Then walk them over to your marketing team.

1. Do all our locations have parity in location-targeting pages?

Most multi-location brands have one flagship location — usually the oldest store, or the largest, or the one with the most attentive marketing team — that has invested heavily in location-specific pages targeting nearby metros. Newer or smaller locations often don't.

In traditional SEO, that gap was tolerable. In generative AI, it's not. The flagship's pages will get cited for the smaller locations' markets.

2. Do our location pages contain explicit geographic signals?

Vague language like *"we serve the [region] area"* is not enough for generative AI to disambiguate which store should win which market. AI needs explicit signals: city names in URL slugs, structured business data (schema markup) with full address, consistent NAP (Name/Address/Phone) across the site, and on-page references to specific cities, counties, and service areas.

The case above is a good test case. The losing pages had vague regional language. The winning pages had the city name literally in the URL.

3. Are we testing on Gemini and Perplexity, not just ChatGPT?

Most brands that look at AI visibility test on ChatGPT because that's where the conversation started. ChatGPT is forgiving. A brand can rank 10 out of 10 on ChatGPT and have a hidden disaster on Gemini and Perplexity that only shows up under direct testing.

Every multi-location brand we've audited that ranks well on ChatGPT has had at least one issue on Gemini or Perplexity. Some of them serious.

What it costs to fix

The fix is mechanical and entirely within your control. It does not require new technology, new agencies, or new ad spend.

For each location that's losing AI citations, the work runs in three steps:

• ●Map the location-targeting page library that the winning location has already built. The pages already exist on a sister store's site — that's the proof of what the AI is willing to cite.
• ●Build matching pages on the losing location's website with explicit geographic signals: city in the URL, local schema, consistent NAP, on-page references to the actual market. Don't reinvent the structure; replicate what already works.
• ●Re-test on Gemini and Perplexity after one content cycle to verify the AI has begun citing the correct location's pages. If the citations have shifted, the gap is closed for those queries.

This is mid-difficulty content work, not a major engineering project. The cost of *not* fixing it scales every month with AI search adoption.

For the framework behind why this works, see our guides on generative engine optimization for dealerships and entity optimization.

A final note for operators

Generative AI search is reorganizing how shoppers find local businesses. The brands that adapt early — by auditing their AI visibility, identifying their own content gaps, and building the missing pages — will own their markets in the AI era. The ones that don't will quietly lose customers to sister stores, to competitors, and to whichever brand happened to publish the right page first.

If you operate a multi-location brand and you're not sure whether AI assistants are recommending the right location to the right shopper, that's exactly the gap we close. The diagnostic takes a single afternoon. The fix takes a content cycle. The compounding loss of doing nothing runs every month.

Pick your three highest-intent buyer queries. Run them on Gemini and Perplexity in each of your markets tonight. Whichever store's URL comes back, that's the store the AI thinks owns that market. If it's not the right one, you've found the page you're missing.