Several people have asked how the system performs a "Near Me" search for food bank near Seattle, WA so I thought it would be a good idea to explain how this is done here on the blog.
As web pages are rolled into the index, the code looks for strings of text that are location references.  In particular, text that is formatted as an address is given a high degree of confidence.  When an address is found, tokens are added to the index that represents the location.  When a “Near Me” search is executed, the engine is told to look for these tokens in addition to the query and adjust the results accordingly. 
The best way to ensure that your site is well represented in these results, assuming your site serves a local need, is to add the address of your business or service on the footer of the page you want selected when a user enters a query that you’re service might be able to satisfy.  Often, sites will place address information on an “about” page.  Unfortunately, these pages often do not rank well with the queries.  Because of potential spam issues, we do not want to attempt to infer the location of a page based on tags in neighboring or child pages.
In order to ensure the address is parsed properly you should format the address as in the following example:

1 Microsoft Way Redmond, WA, 98052

Notice the capitalization for the city name and state.  Also, make sure the city and state are not separated by any html tags. Following the above simple suggestion is easy and will help users find you with our system.
When a “Near Me” query enters the system  uses  the user’s ip  address to guess their location via a process called ReverseIP lookup. This data is stored in a non-identifying cookie for future searches. It can be changed  at any time by going to the Settings page.

UPDATE: For the time being only US adresses will work while we develop this feature.  Over time we will be adding more markets.  We will post to the blog as we add new markets.

-Chris Weare, MSN Search Developer