SEO, AEO and GEO in the US: winning metro markets, not national keywords

Short answer: The US is not one search market. It is hundreds of metro markets under a growing patchwork of state privacy laws, and national keywords are the wrong target for almost every small business. Win the metro, get granular only where you have real substance, and build the review and licensing signals that American AI answers actually draw on.

Most search advice is written as though a country is a single market with a single set of rules. In the United States that assumption fails twice over: once on geography, where a business competes in a metro rather than a nation, and once on regulation, where the applicable privacy rules now depend on which state your visitor is sitting in.

For the plain definitions of the three disciplines, SEO, AEO and GEO explained covers them. What follows is what changes when you apply them here.

The scale trap

National keywords look like the prize and are usually the trap.

A broad service term in the US is contested by companies whose marketing budgets exceed most small businesses’ annual revenue. Worse, if you win any of it, most of the traffic comes from places you cannot serve. You pay in effort and content for visitors who were never addressable.

Metro-level intent inverts both problems. Competition is a fraction of the national field, the traffic is inside your service area by definition, and the buyer is closer to a decision because they searched with location intent in the first place.

The practical rule: national terms are for national products, and even then you enter through a specific segment rather than the head term. Everything else starts at the metro.

Metros behave like separate countries

Treating “the US market” as one plan produces content that fits nowhere in particular.

Competitive intensity varies enormously. Winning a service term in a top-ten metro is a different undertaking from winning it in a mid-sized city, with different timelines and budgets. Vocabulary shifts regionally, and the same service is described differently across the country. Pricing expectations differ so widely that a single national price page reads as wrong in both directions at once, too expensive in one market and implausibly cheap in another.

Seasonality diverges too. A campaign timed for a single spring does not describe a country where the practical season for outdoor trades starts months apart across regions. What digital marketing and advertising actually cost in San Antonio is a case in point: Fiesta draws crowds every April, so a hospitality or events business there plans around its own local peak, one that a statewide Texas calendar would miss.

Local SEO across Sacramento’s four county markets shows this in miniature, with downtown, Folsom, and Elk Grove each running their own map pack contest.

The workable structure is a national core that establishes what you do, with metro pages carrying the specifics that actually differ.

The state privacy patchwork

There is no single federal privacy law governing this, and the gap has been filled state by state.

California’s CCPA and CPRA set the strictest broadly-applicable requirements, and a growing number of states have followed with their own laws at differing thresholds. The result is a compliance surface that depends on where your visitor is, not where you are.

The important clarification: none of this restricts what you publish. Privacy law does not limit SEO, AEO or GEO content, and treating it as a reason to publish less misreads it entirely. What it governs is the capture layer, meaning consent handling, tracking, disclosures, and honoring deletion and access requests.

That connects to visibility indirectly but genuinely. A tracking setup that ignores consent creates exposure on precisely the traffic this work generates, and a form that over-collects turns a marketing asset into a liability.

The ADA lawsuit exposure

State privacy law is not the only compliance question a US marketing site has to reckon with. A separate and faster-moving one is web accessibility litigation under Title III of the Americans with Disabilities Act, which has nothing to do with what you publish and everything to do with whether every visitor, including one using a screen reader, can actually use the page.

Federal website accessibility lawsuits reached 3,117 in 2025, up from 2,452 in 2024, a 27 percent increase that Seyfarth Shaw’s ADA Title III tracker describes as recovering ground lost to declines in the two prior years. Title III’s own language carries no revenue or headcount threshold that exempts a smaller business, and a company does not need to be a household name to receive a demand letter over an inaccessible page.

None of this changes what to write. It changes how the page is built: alt text on images, keyboard navigation, sufficient color contrast, and a structure a screen reader can parse in the order a sighted visitor would read it. A site built to a reasonable accessibility standard from the start is also, not coincidentally, easier for an AI crawler to parse cleanly, which makes this one of the few compliance obligations that pays back in visibility rather than only reducing risk.

Bing is not a rounding error here

Google dominates, and Bing still holds a meaningfully larger share in the US than in most comparable markets. It also feeds Microsoft Copilot.

This matters most for B2B, where a substantial amount of traffic originates in managed corporate environments running Edge with Bing as the default. Businesses selling to enterprises routinely discover they have been invisible to a slice of their actual buyers.

The practical response is small: verify Bing Webmaster Tools, submit the sitemap, and confirm indexing. The same clearly structured content serves both engines, so this is a setup task rather than a separate program.

The American trust stack

Review platforms carry unusual weight in US consumer categories, and this feeds both conversion and citation. Yelp retains real influence in several verticals, Google reviews are close to universal, and in regulated trades, state licensing databases function as verifiable proof in a way marketing copy never can.

Two consequences follow for AI visibility. First, review text matters more than review score. A business with detailed reviews gets described in specific terms when an AI is asked what it is like to work with; a business with a higher average and no text gets described vaguely or not at all. Second, licensing and accreditation records are exactly the corroborating sources generative systems favor, because they are structured, public and verifiable.

Getting reviews consistently is the fastest lever most US local businesses have. It is also the one most reliably neglected. Local SEO for US small business covers how to ask for them within the FTC’s review rule and Google’s own policy.

Local Services Ads and the Google Verified badge

For service categories that qualify, Google Local Services Ads sit above both the paid and organic results on a local search, carrying a Google Verified badge (which replaced the older Google Guaranteed and Google Screened badges in 2025), a star rating, and a review count before a searcher sees anything else. Unlike a standard Google Ads campaign, this format charges per lead rather than per click: a phone call or message that reaches the business, not an impression or a visit to the site.

SearchLight Digital’s February 2026 benchmark, tracking US$6.72 million in Local Services Ads spend and 126,650 leads across 888 home service contractors, put the average cost per lead at US$53. By trade, electrical ran the lowest at US$39, HVAC US$51, plumbing US$57, and drain and sewer work the highest of the tracked categories at US$59. A business can dispute a lead that turns out to be a wrong number or outside its service area, though not every disputed lead gets credited.

The badge itself functions as another entry in the trust stack rather than a separate program. Google verifies the business behind it through a background check and license review before granting it, which makes it a structured, machine-checkable trust signal in the same category as a licensing record, the kind of fact both a searcher and an AI answer engine can rely on without having to interpret marketing copy.

Building for neighborhood intent

In a large metro, intent gets granular fast. People search by neighborhood, by suburb, sometimes effectively by ZIP code.

That granularity is worth pursuing exactly as far as you have real substance. A page about serving a specific area should contain things only someone who works there knows: travel times, permit or inspection differences, the housing stock or business types, work actually performed nearby.

Where that substance runs out, stop. A page that changes only the place name is a doorway page, Google acts against the pattern, and generative engines will not cite it because there is nothing specific in it to quote. Ten substantive area pages beat sixty templated ones, and carry none of the risk.

Precision beats scale in local search

Two patterns show up repeatedly in city-level work that a national or even a metro-level plan misses.

The first is name collision. San Jose is not only Silicon Valley’s biggest city, it shares its name with the capital of Costa Rica, one letter apart on airport codes that travel agents and travelers still mix up regularly. A business that never states its city and state in full is leaving an AI system to resolve two real cities with the same name, which is a solvable problem at zero cost: say San Jose, California, not just San Jose, on the pages that matter.

The second is demand that moves. Bellingham’s retail and hospitality economy drew roughly 12 percent of its taxable sales from Canadians crossing the border to shop, until that traffic fell by close to a third in 2025 and has not recovered. No amount of on-page optimization brings that specific demand back. What a local SEO and GEO plan can do is redirect toward the demand that is still there, whether that is university students, ferry passengers, or a seasonal event, faster than a paid media plan can be re-budgeted.

Both cases say the same thing about granularity: the city page has to reflect what is actually true about the city right now, not a template filled in with a place name.

Multi-location and franchise listings

A franchise or multi-location business hits a different setup problem than a single-location one, and it shows up first on Google Business Profile.

Google offers bulk verification to businesses with ten or more genuine brick-and-mortar locations, grouped together as a location group inside Business Profile Manager rather than verified one at a time. Service-area businesses without a public storefront do not qualify. An authorized representative attests to the relationship between the locations and the parent brand, and Google’s own help documentation states that new locations added to the dashboard afterward, if not already claimed by someone else, show directly on Google rather than each waiting on a separate verification request.

Bing Places for Business solves the same problem on a different model: one account manages every location, and a spreadsheet upload can create or update a large batch of listings at once instead of logging into each one individually. For a franchise running dozens or hundreds of units, that is the difference between weeks of manual entry and an afternoon.

The harder part is not verification, it is consistency once every location is live. Name, address and phone number need to match exactly across the parent site, each location’s own listing, and any franchise directory, because search engines and AI systems use that match to decide whether they are looking at one business with many locations or several unrelated ones. A location page carries the same requirement. It should state the specific address, hours and service area for that location rather than swap a city name into an otherwise identical template, because a page that changes only the place name reads as a doorway page and gets treated as one regardless of how large the brand behind it is. Corporate can set the standard, but a franchisee’s day-to-day listing upkeep is often looser, and inconsistency creeps in fastest at the individual location, not at headquarters.

Where to start in a crowded market

Reviews first. Fastest lever, largest effect on both local ranking and how AI answers describe you, and it costs nothing but a process for asking, one that AI marketing tools can automate as long as no one drafts the review itself.

Then the entity. Consistent name, address and phone across your site, Google Business Profile and the major directories, plus licensing records where they apply. Inconsistency here is what stops a citation.

Then answers. The questions you get asked on the phone, written plainly on your own pages. This produces visible movement in weeks while ranking work is still developing.

Then metro depth. Real pages for the areas you genuinely serve, built only as far as your substance goes.

National terms last, if ever. For most businesses reading this, never is the correct answer.

Budgeting in dollars

The cost driver in the US is competitive intensity, and it varies by an order of magnitude between a top-ten metro and a mid-sized city.

Technical and entity work is a modest one-off almost anywhere. Content scales with how contested your metro is, because the bar for what is good enough is set by whoever already ranks. Reviews cost process rather than money. For what a monthly quote buys in hours, and when a hire beats a retainer, see how much SEO costs in the US.

Expect three to six months in a mid-sized metro and six to twelve in a major one for commercially valuable local terms, with answer-shaped content producing citations considerably sooner. If you need results faster than that, paid search is the honest answer, and we will tell you so rather than sell a program that cannot deliver in your timeframe. Our guide to digital advertising for US small business covers that route.

AI Overviews in American results

Google began rolling AI Overviews out to US searchers the week of May 20, 2024, following the announcement at Google I/O on May 14 that year, and the feature has since become a fixture of American search rather than a temporary experiment.

The clearest independent read on what that means for clicks comes from the Pew Research Center, which tracked the actual browsing activity of 900 US adults across roughly 70,000 Google searches in March 2025. Fifty-eight percent of those adults saw at least one AI-generated summary that month. On searches where a summary appeared, users clicked through to a traditional web result only 8 percent of the time, against 15 percent when no summary showed. Clicking a link inside the summary itself happened on just 1 percent of visits, and people were more likely to end their search without clicking anything at all when a summary was present.

For a US business, the practical shift is where the work has to pay off. A page can hold its ranking position and still produce roughly half the clicks it used to on the same query, because the summary answered the question before the searcher reached a results list. That makes being the source quoted inside the summary, not just the top link below it, the thing worth chasing. The pages most likely to get pulled into an Overview are the ones stating a plain, specific answer near the top, the same discipline that wins featured snippets and AI citations generally. Our AEO checklist for US small businesses sets out the eight jobs in order.

It also changes what a falling click count means. A page losing clicks while holding position may be getting summarized rather than failing, and the way to tell is to check whether it still appears as a citation inside the Overview. For the assistants outside Google, see why ChatGPT may not be recommending your business.

What to track when volume is high

US volumes are large enough that weekly numbers carry real signal, which is a genuine advantage over smaller markets.

Segment by metro rather than nationally, because a blended figure hides the market where you are actually competing. Track enquiries and cost per enquiry alongside rankings, since position without contact means the snippet or the offer is failing rather than the ranking.

Then test the AI answers directly, metro by metro. Build the prompts a local buyer would use, run them on a schedule, and record who gets named instead of you. In US local categories the answer is often a review platform or a national aggregator, which tells you exactly where the citation work needs to happen.

Where US businesses waste the most

• Chasing national terms. Expensive, slow, and most of the traffic is unservable.
• Templated neighborhood pages. A doorway pattern with real penalty risk and no citation value.
• Neglecting reviews. The fastest available lever, left idle.
• Assuming Google is the whole market. Particularly costly in B2B.
• One national price page. Reads as wrong in most of the country simultaneously.

Frequently asked questions

Should a US small business target national keywords at all? Almost never, at least not first. National head terms in the US are contested by companies with marketing budgets larger than most small businesses’ revenue, and the traffic they deliver is mostly outside any service area you can actually serve. Metro and neighborhood-level intent converts far better and costs far less to win. The exception is a genuinely national product with no geographic constraint, and even then the sensible entry point is a specific segment rather than the broad term.

How much do state privacy laws affect marketing sites? More every year, and unevenly. California’s CCPA and CPRA set the strictest widely-applicable baseline, and a growing number of states have passed their own laws with differing thresholds and requirements. The practical effect on a marketing site is on consent handling, tracking, and how you disclose and honor data requests. None of it restricts what you publish, so it does not limit SEO or AI visibility content. It shapes the forms and tracking behind that content.

Does Bing matter in the US, or is it all Google? Google still dominates, but Bing holds a meaningfully larger share in the US than in most other English-speaking markets, and it feeds Microsoft Copilot. For B2B especially, where a lot of traffic comes from managed Windows environments running Edge with Bing as default, ignoring it leaves real volume unclaimed. Practically, the same structured, clearly written content serves both, so this is a reason to verify Bing Webmaster Tools rather than to run a separate program.

How granular should local pages get in a large metro? As granular as you have genuine substance for, and no further. In a large metro, neighborhood-level intent is real and worth targeting, but only where you can say something specific about serving that area: travel times, permit differences, the kinds of properties or businesses there, actual work performed. A page that changes only the neighborhood name is a doorway page, and Google acts against those. Ten substantive area pages outperform sixty templated ones.

What do US AI answers cite most often? Review platforms with real volume, established trade and local press, government and licensing databases, and businesses whose own pages state verifiable facts plainly. Yelp and Google reviews carry unusual weight in US consumer categories, and licensing or accreditation records matter in regulated trades. A business with detailed local reviews gets described in richer terms by an AI answer than one with a higher rating and no review text.

How long does this take in a competitive US metro? Answer-focused content can produce citations within one to two months. Local ranking in a mid-sized metro typically takes three to six months. In a top-ten metro against established competitors, expect six to twelve for commercially valuable terms. The variable that moves this most is review volume and recency, which is the fastest lever most US local businesses have and the one most consistently neglected.

How much have AI Overviews actually cut organic clicks in the US? A lot, based on independent data. The Pew Research Center tracked real browsing behavior across roughly 70,000 US Google searches in 2025 and found people clicked a traditional result in just 8 percent of searches where an AI summary appeared, against 15 percent when none did. The response is not to abandon SEO, it is to write the specific, quotable answer that gets pulled into the summary itself.

Should a US service business use Google Local Services Ads? For qualifying trades, yes, and it functions differently from standard PPC. Local Services Ads charge per lead rather than per click, averaging US$53 in SearchLight Digital’s February 2026 benchmark across 888 contractors, and display the Google Verified badge above both organic and paid results. That badge is a verified trust signal in the same category as a licensing record, useful to a searcher and to an AI system deciding who to cite.

Is ADA website compliance a real legal risk for a small US business? Yes, and it is a faster-moving risk than most state privacy law. Federal website accessibility lawsuits reached 3,117 in 2025, up from 2,452 in 2024, a 27 percent increase. Title III’s own language carries no revenue or headcount threshold that exempts a smaller business, so size alone is not protection.

Why would an AI system need my state as well as my city name? Because some US city names are not unique, and a few are not even unique globally. San Jose, California shares its name with the capital of Costa Rica, and their airport codes, SJC and SJO, are one letter apart. Stating your city and state in full on the pages that matter costs nothing and resolves an ambiguity an AI system otherwise has to guess at.

---

Want to see what AI answers say about your business in your metro? Get a free SEO and AI visibility audit. We will show you who is being named instead of you.