Local schema for multi-location trades businesses

A trades business with more than one location usually has one website with a locations page, one block of structured data describing the parent company, and two shops quietly competing for the same searches without either one ranking well. The site is telling a search engine that there is a single business here. The business is telling its customers there are two.

What the search engine actually needs

A distinct entity per location. Each shop needs its own page, its own address, its own hours, its own phone number, and its own structured data block describing that location rather than the brand. Without that, there is nothing to attach a local ranking to except the domain, and the domain can only rank once.

The practical test is simple. If you take the schema block from your Wilmington page and drop it onto your Hockessin page, does anything change? If the answer is no, you do not have location schema. You have company schema on two URLs.

The data model matters more than the markup

The markup is the easy part. What breaks multi-location sites is that the location data lives in six places: the header, the footer, the locations page, the contact page, the schema block, and the sitemap. When a phone number changes, five of them get updated and one does not, and the one that does not is usually the schema, because it is invisible.

The fix is structural. Put every location in one typed data file and generate everything from it: the routes, the navigation, the sitemap entries, the schema blocks, the booking links, and the call to action. Then a new address is one edit, and there is no second place to forget.

That also makes the third location cheap, which matters because a business that opened a second location is usually planning a third.

Do not point every location at the same phone number

If all locations route to one number, call tracking cannot tell you which location produced the call, and neither can the search engine infer that these are genuinely separate places of business. Separate numbers per location are worth the small operational annoyance.

The same applies to hours. Locations that share a single hours block are usually wrong for at least one of them, and hours are one of the fields a search engine surfaces directly.

What to expect

Splitting locations properly does not produce an overnight jump. What it produces is the ability to rank at all in the second location's town, which a single blended entity cannot do. The change tends to show first in map pack impressions for the smaller location, because that is where the previous setup was costing the most.

Where this fits

This is the structural half of local search. The content half is service pages that say something different per location rather than the same paragraph with the town name substituted, which search engines have been discounting for years and which readers notice immediately.