A business serving several markets should run one domain with a clear location hierarchy, not several websites. One site concentrates every link, every review mention, and every bit of authority you earn; several sites split all of it and triple the maintenance. The harder question is not one site or many, but how many city pages you can genuinely justify, and this guide covers the multi-location website structure that scales without producing filler.
Key takeaways
- One domain. Separate sites per city split your authority and multiply your work for no ranking benefit.
- Put locations in a folder, not at the root, so the structure stays legible as you add markets.
- Write only the city pages you can make genuinely different. A page that would still be true with the name swapped is a liability.
- Do not build a service-by-city grid unless each combination has real demand behind it.
- Physical locations and service areas are different things, and Google treats them differently.
Should I build one website or one per location?
One, unless the locations are genuinely separate businesses with separate brands and separate ownership.
The argument for separate sites usually comes from a reasonable instinct: a dedicated site for each market feels more local. In practice it costs you three ways.
- Authority splits. Every link, mention, and review someone gives you lands on one of five sites instead of one, so none of them builds much weight.
- Maintenance multiplies. Five sites means five sets of updates, five security surfaces, five hosting bills, and five things to forget.
- Content competes. Five near-identical roofing pages on five domains owned by the same business is a pattern search engines recognise and discount.
The exception that genuinely justifies separate sites: distinct legal entities operating under different brand names, where a visitor to one would be confused to land on the other. Franchise arrangements sometimes force this. A single company serving five counties does not.
Where should location pages live in the URL structure?
In a dedicated folder, with the hub above the individual pages.
/locations/ the hub, listing every market
/locations/lakeland/ an individual market
/locations/winter-haven/
/services/ what you do, independent of where
/services/roof-replacement/
This does two useful things. It keeps the top level of the site readable as it grows, and it makes the relationship between pages obvious to both visitors and crawlers: these pages are siblings, and they belong to that hub.
The alternative you will see is city pages at the root, like /lakeland-roofing. It works for a handful of pages and becomes unmanageable at twenty. There is no ranking penalty for the folder, and the folder is easier to live with. On our own site we group these under /web-design for exactly that reason, with county hubs above the city pages.
Two structures to avoid:
- Nesting more than two levels deep (/locations/florida/polk-county/lakeland/roofing). Every level adds distance from the home page and clicks for the visitor.
- Duplicating a full service tree inside each city folder, which produces the service-by-city grid problem below.
How many city pages should I actually build?
As many as you can write genuinely differently, for places you genuinely want work from. That is usually a much smaller number than the list of places you technically serve.
A useful filter, applied in order:
- Do you want more work here? Not "would you accept it." A city an hour away where the drive kills the margin does not need a page.
- Can you say something specific? Named neighbourhoods, local building stock, permitting quirks, weather patterns, projects you have done there.
- Is there search demand? Some small towns have almost none, and the page will never earn its maintenance.
- Can you keep it current? Twenty pages you never revisit age badly.
Ten strong city pages outperform sixty thin ones, and the thin ones are not neutral. A large volume of near-duplicate pages is a signal that dilutes the pages you actually care about, and it is one of the most common self-inflicted local SEO problems.
What makes a city page substantial rather than thin?
The test: would this page still be true if you swapped the city name? If yes, it is thin.
| Thin | Substantial |
|---|---|
| "We proudly serve Lakeland with quality service" | Named neighbourhoods and the housing stock you work on there |
| The same three service blurbs as every other city page | The services that actually sell in that market, in order |
| A stock photo of a generic house | A photo of your own job in that city |
| "Contact us today for a free quote" | The local phone number, and how far out you are booking |
| No mention of anything local | Permitting, HOA norms, climate, or ground conditions specific to the area |
Local specifics are the part owners find hard, because they know it and have never had to write it down. A productive shortcut: record yourself answering "what is different about working in this town" for two minutes, and write from that. It is invariably more specific than anything a writer could invent.
Reviews from customers in that city are the other strong signal. Around nine in ten customers read reviews before choosing a local business, and reviews that mention the place name do double duty. If review collection is ad hoc, automating the ask is a higher-leverage fix than another page.
Should I build a page for every service in every city?
Only for the combinations with real demand and real margin behind them.
The arithmetic explains why this goes wrong. Eight services across twelve cities is 96 pages. Almost nobody has 96 pages' worth of genuinely distinct things to say, so the pages get generated from a template, and the result is 96 near-duplicates that compete with each other and with the eight real service pages.
A saner approach: build the service pages properly, build the city pages properly, link them to each other, and add a service-by-city page only where you can point at actual search volume and actually want that specific work. That is usually a handful of combinations, not a grid.
How do physical locations differ from service areas?
Google treats them differently, and getting it wrong affects your Business Profile more than your website.
- A physical location is an address customers can visit, with staff and normal opening hours. It gets its own Google Business Profile and warrants its own page with the address, hours, parking, and a map.
- A service area is where you travel to work from a single base. That is one profile with a service area defined, not one profile per city you drive to.
Creating profiles for addresses you do not genuinely occupy is a violation of Google's guidelines and gets listings suspended, which is a far worse outcome than ranking modestly. The rules are set out in Google Business Profile Help, and they are enforced.
The website side follows the same logic: a page for a real branch can talk about visiting, parking, and hours. A page for a service-area city should talk about the work, not pretend to an address.
How should the location pages link together?
Every city page links up to the hub and out to the services it covers. The hub links down to every city. Service pages link to the markets where that service is offered.
That sounds obvious and is frequently missed, usually because city pages are generated in a batch and never linked from anywhere except a sitemap. A page reachable only from the sitemap is a page search engines will treat as unimportant, because that is what your own site is telling them.
Two additional habits that help:
- Cross-link neighbouring cities from each city page. It is genuinely useful to a visitor who searched the wrong town, and it distributes internal links.
- Link from your guides. An article about roofing costs that links to the markets you serve gives those pages a route in from content that attracts links of its own.
How do I add a new market later?
Treat it as a small launch, not a copy-paste.
- Confirm you want the work and can service it profitably.
- Write the page from real local specifics before it goes live, not after.
- Link it from the hub, from neighbouring city pages, and from the relevant service pages.
- Update the Business Profile service area if the coverage genuinely changed.
- Check it renders and is indexable, using the same launch checks you would use for any new page.
The failure pattern is adding ten markets in an afternoon because the template made it easy. Adding one properly every month or two beats adding ten badly in one sitting.
What to do next
List every place you currently claim to serve, then cross out every one you do not actively want more work from. What remains is your real location list, and it is usually a third of the original.
If the structure needs rebuilding rather than pruning, that is the kind of thing we plan before we build. Tell us which markets matter and we will map it out with you.
