The Problems

Search Engines can’t fill out a store locator box.  Google-bot can’t type a zip code.  So even though users can easily use these kind of interfaces to find their nearest store location, a search engine can not do that.

As a result, a directory of store locations had to be developed as a path that Search Engines can use to find the store pages.  This was effective until one day in May.  The An SEO Case Study in Getting Store Locations Found in Natural Search directory structure could only be accessed from one link on the site, deep in the site map page.  When the Google May Day update hit, the directory structure had all of the store locations 4 clicks away from the home page.  Ever since the May Day update you really need to try to have your content within 2 clicks of your Home Page.

The client also used JavaScript to dynamically pull the addresses, phone numbers and store names from a database and then put that content in JavaScript on the store location pages.  Search Engines can’t really read JavaScript content.  So if a Search Engine somehow did bother to index all the way down to the store location page level (which they didn’t) their wasn’t any readable content with regard to the location of the stores.

The client also had a couple of different websites, which also used the same exact store location pages just with different URLs.  This is duplicate content in the eyes of Search Engines.  Chances are a Search Engine will usually only rank one of the duplicate content URLs, and it may not be the URL that you prefer to rank.

At its worst only 10% of the over 600 location pages had top 10 rankings for our selected keywords which were a variation of the town name and the store names on Google.  Before the May Day update and the use of JavaScript to produce the Name, Address and Phone Number well over 75% of those words had top ten rankings for those same keywords.

The Solutions

Search Mojo cleaned up the directory structure of the store locations and convinced the client to add a footer link to all pages of the website to that directory structure.  This had a significant impact on improving many keyword rankings.  It led to the store pages at least getting found and indexed by Google, which wasn’t happening with the directory structure being buried deep within the site, far from the Home page.

The client was unwilling to abandon the use of JavaScript content on the store pages for the Names, Addresses and Phone Numbers.  However, we convinced them to utilize “noscript tags” that are only visible to Searchers that do not have JavaScript enabled on their browsers.  The identical content for the Names, Addresses and Phone numbers was put in HTML within those noscript tags.  This gave the search engines something they could read containing the name of the store and the town where those stores are located.

Canonical tags which refer search engines to rank the pages that you want ranking rather than their duplicate content counter points on other URLs were implemented to avoid issues of duplicate content.

The Results

Rankings bounced back for the store location pages.  We went from 10% of them having top ten search rankings on Google to about 70% of them having top ten rankings on Google.  Many other pages are very close to becoming top 10 ranked pages.

What Could Have Worked Better

The HTML content within the noscript tags was a good band aid that was useful in improving the selected keyword rankings of store pages.  However, our preference was to abandon the use of JavaScript for this content and use HTML and Rich Snippets.  Google has specifically said that use of Rich Snippets within noscript tags will not work.

Rich Snippets have the potential to give search engines the ability to associate your page with a location, which can be a big deal.  When search engines can personalize search results based on the searchers location (especially on mobile devices enabled with GPS) these store pages can lead to increased store visits and sales.