Every scaling service business hits the same ceiling. You own your home city (map pack, organic results, phone rings) and then you look at the county next door, where your trucks already drive three days a week, and wonder why you're invisible there.
So someone builds five location pages–same 600 words on each, city name swapped, maybe a stock photo of a skyline. Six months later: no rankings, no calls, and a site Google crawls with noticeably less enthusiasm than before.
That isn't bad luck, and it usually isn't a penalty in the way people mean the word. It's architecture. Google doesn't punish you for serving ten cities, but it does ignore you. Why? You didn't introduce any new information for nine of them.
Dropping a pin on a map is not a strategy. A real multi-location local SEO strategy is a structural system: a location hierarchy search engines can crawl, content with genuine local substance, structured data that states plainly where you work, and answer-ready formatting that gets you named by AI search. Here's the system that works.
Know Which Business You Are
It's critical that you get this right before you build a single page, because every downstream decision depends on it. A brick-and-mortar business—dealership, dental office, retail counter—serves customers at a staffed address during posted hours and ranks outward from it. A service-area business (SAB)—HVAC, plumbing, roofing, moving and logistics, mobile automotive—travels to the customer, and Google requires SABs to hide the street address and list up to 20 service areas instead.
So what blindsides operators? It's the fact that listing 20 service areas does not make you rank in 20 cities. Your profile still ranks from your verified address outward, and proximity remains one of the strongest signals in the map pack. As much as we'd like for it to, no amount of service-area configuration overrides physical distance.
This means the map pack in a city 40 minutes away is effectively closed to you without a real, staffed location there. What stays wide open is everything below the map—organic results, and the AI summary sitting above them. That is the ground your location pages are actually fighting for, and it's worth more than most operators assume.
Renting a virtual office, a mailbox store address, or using a relative's garage to "plant a flag" in a target city is the fastest way to lose the profile you already earned. Google's suspension systems are good at spotting shared addresses and unstaffed locations, competitors report them routinely, and reinstatement can take weeks, if it comes at all. Never trade a ranking profile for a fake one.
URL Structure That Scales
Two patterns are defensible, and the only wrong move is using both: location-first (/service-areas/city-name/ as the city hub, with /service-areas/city-name/service-name/ underneath for high-value city × service combinations) or service-first (/services/service-name/city-name/). Location-first is the better default for most service-area businesses, because the city hub becomes a genuine landing page for "near me" searches and a natural home for local proof, reviews, and team.
Whichever you choose, apply it sitewide—running both creates two URLs chasing the same query, and you end up competing with yourself. Three rules hold either way:
- Stay shallow. Hub one level off the root, spokes one level below that. Anything deeper gets crawled less and converts worse.
- Be consistent with trailing slashes across links, canonicals, and your sitemap. Inconsistency creates duplicate URLs out of thin air.
- Link deliberately. Each hub links down to its spokes, each spoke links back to its hub, and neighbouring cities link to each other. Do not dump forty city links into your footer—that tactic served us best back in 2014.
The number of city pages you should build equals the number of cities you can prove you work in: named jobs, photos, reviews, a crew that covers it. Not the number of cities you'd like to work in. Build in waves of three to five as the proof accumulates, and you'll never publish a thin page.
The Thin Content Trap
Find-and-replace location pages fail for a mechanical reason. When twelve pages share 95% of their text, search engines treat them as near-duplicates, pick one to represent the set, and quietly drop the rest. Worse, site-wide quality signals mean a mass of low-value pages can drag on the pages that were performing fine. The cost of a bad location page is negative (no, not even zero).
Use this test: cover the city name with your thumb. If the page could describe any market in the country, it isn't a location page. It's a template with a variable in it, and search engines read it exactly that way.
Anatomy of a City Page That Earns the Call
A location page that ranks and converts is doing four jobs at once.
- Local proof, not local keywords. Repeating "plumber in Sartell" eleven times does nothing; evidence that you work in Sartell does everything. Three named jobs with the neighbourhood, the problem, and the fix beat four hundred words of filler—then add the local reality (hard water, ice dams, 1960s cast iron, the permit office you pull from), which crew covers the area, and photos of your team on a job there. Stock and AI imagery lose every time.
- Map, NAP, and location schema. Embed a map of the area you actually serve and show the full name, address, and phone exactly as they appear everywhere else. No staffed location in that city? Show the service radius—never invent an address to make a page look local.
- Reviews and case studies tied to that geography. Surface the reviews that name the city on that city's page, and anchor each hub with at least one full case study from the area. It gives AI systems concrete, locality-linked evidence to cite.
- Localized service list and a two-speed CTA. Lead with what that city actually buys, then offer one path for people ready now (call, emergency dispatch) and one for people comparing (online booking, estimate request). Forcing a phone call on a homeowner researching a roof replacement at 10pm costs you the lead.
The Technical Layer: Schema, Canonicals, and AEO
Local schema markup for contractors is where most sites leave easy ground uncovered. Use the most specific type available—HVACBusiness, Plumber, RoofingContractor, MovingCompany—then let three properties do the heavy lifting: areaServed, which explicitly declares your footprint; hasOfferCatalog, which tells a machine the specific menu you offer in that market; and sameAs, pointing at your Google Business Profile and industry listings, because corroboration builds entity confidence. One caution: emit one LocalBusiness node per real location, not one per page. Stamping every city page with an address you don't occupy is a structured-data misrepresentation, and it's exactly the signal that invites a manual review.
On canonicals, every city page should self-canonicalize, pointing at its own absolute URL, trailing slash included. The costly mistake is canonicalizing all your city pages to the main service page on the advice that it "avoids duplicate content." What that actually does is tell Google to consolidate those pages away and index none of them. You built twelve pages and asked search engines to ignore all twelve. Canonicals are for true duplicates like tracking parameters and print views; if two pages genuinely cannibalize each other, merge the content and 301 the weaker URL.
Then there's AEO. When someone asks ChatGPT, Perplexity, or Google's AI Overviews "who is the top-rated HVAC contractor in [City]?", the system is matching an entity to a service to a location, and it strongly favours content that makes all three connections in plain language. So ask the question the way a customer would as an H3, answer it in the first 40–60 words, and keep every answer self-contained enough to survive being lifted out of context. Add FAQPage schema with three to five genuine local questions, and keep your name, address, and phone identical everywhere—conflicting data makes an AI system less confident recommending you, and low confidence means it names a competitor instead. If this is new territory, our breakdown of what AEO is and how to implement it covers the fundamentals in more depth.
Off-Page: Signals With a Zip Code Attached
One Google Business Profile per genuinely staffed location—real signage, real people, staffed during the hours you post. Never keyword-stuff the business name: "Smith Plumbing" is compliant, "Smith Plumbing | Emergency Plumber Minneapolis" is the single most-reported and most-suspended violation in local search. And point each profile at its own city page, not the homepage, because that link is a relevance signal and wasting it on the homepage wastes it entirely.
Then earn links with a zip code attached: chamber of commerce memberships in each county you actively serve, manufacturer and supplier dealer locators, sponsorships with a webpage attached, local news coverage and trade association chapters. Five real links from organizations in a target city outrank a hundred generic directory listings. Chase the five.
Reviews do double duty—ranking signal and quotable evidence for AI answers. Ask at peak satisfaction (job completion, then again by text within 24 hours with a direct link), prompt naturally rather than from a script, route each request to the nearest profile, reply to every one, and feed them back onto the city pages so each review works twice. Our guide to getting more Google reviews the right way covers the compliance side in detail.
What This Looks Like in Practice
Here's the sequence, using a composite of the multi-market builds we run—a residential HVAC operator with one office, twelve trucks, and technicians already working across five surrounding counties. Rank markets by actual revenue from two years of job data, not ambition. Cut twelve existing thin city pages down to the five markets with real job volume and redirect the rest. Build those five hubs on proof gathered in a half-day of ride-alongs, then wire up the technical layer: self-canonicals, LocalBusiness and FAQPage schema, hub-to-spoke internal links, a sitemap cleaned to canonical URLs only. Add chamber memberships in the two highest-value counties, dealer-locator listings, and a review workflow that routes by job address. City × service spokes come in month four, and only for combinations with proven demand and margin.
What changes isn't a single ranking. It's that organic calls from outside the home city stop being incidental and become a predictable share of the pipeline—which in turn lowers the pressure to bid on every satellite city in Google Ads just to stay visible. Expect early movement in 60–90 days and meaningful compounding by month six. Anyone promising faster is selling something.
Three Things That Actually Matter
- Architecture. One URL pattern, shallow hierarchy, deliberate internal links, self-canonicals. Structure determines whether anything else you do can work.
- Unique content. Real proof from real jobs in each market. If the page passes the thumb test, it isn't finished.
- Local signals. Compliant profiles, links from organizations in that city, and a steady flow of reviews that name the place and the work.
Do those three consistently and you can compete across ten cities from one location—without a fake address, and without a site full of pages search engines have already decided to ignore.
If you want to know where your current setup stands, book a free digital audit with our team. We'll review your location architecture, schema, profiles, and review signals, and give you a prioritized roadmap for the markets worth expanding into next.
Frequently Asked Questions
How many location pages should I build?
Build one page per city where you can prove you work — named jobs, photos, reviews, and a crew that covers it. Ten well-evidenced city pages outperform forty thin ones, because near-duplicate pages get filtered out of the index and can drag on site-wide quality signals. Add pages in waves of three to five as the proof accumulates.
Should I canonical my city pages to my main service page?
No. That tells Google to consolidate those pages away and index none of them. Every city page should self-canonicalize to its own absolute URL, trailing slash included. Canonical tags are for true duplicates like tracking parameters — if two pages genuinely cannibalize each other, merge the content and 301 the weaker URL instead.
Do I need a separate Google Business Profile for every city I serve?
Only for cities with a genuinely staffed location that has real signage and posted hours. Service-area businesses use one profile with the address hidden and up to 20 service areas listed. Renting a virtual office or mailbox address to create profiles in other cities is the most common cause of suspension, and reinstatement can take weeks.
How long do location pages take to rank?
Expect early movement in 60–90 days and meaningful compounding results by month six. Pages backed by real local proof, correct schema, and location-specific links move faster than templated pages, which often never rank at all regardless of how long you wait.