The folks over at RankRanger published data around a local pack study that they are calling the 2:1 pattern. I believe, if I understand the interpretation of the data properly, that they are saying that the local pack, which contains three results, shows two of those results with really local and nearby businesses and the another business that is not as near as the others.
Google does this provide two really close and local results (relative to the search location), and then one further away for a bit of diversity, is how Joy Hawkins explained the study.
He reviewed 131 Local Packs and came up with that "Google is clustering two of the locations within the Local Pack together and then offering a third, more distant location, or 2:1."
It is an interesting study, because personally, in my several queries that I intentionally brought up local packs for my specific area (i.e. pizza, haircut, accountants, etc), didn't show me this 2:1 pattern. But I recommend you look at all the data they shared over here - it is pretty interesting.
Here are their "why it matters:"
- The showing of the 2:1 pattern on the state level is odd, since the nature of the query does not intend to be limited to one location (i.e. in Texas Google had a propensity to show two clustered listings near Houston).
- It points to the algorithm behind the Local Pack perhaps being too focused on search location. In other words, Google is treating the user from around the corner and the tourist from around the world the same way.
- The 2:1 pattern also has implications when the location of the business is secondary (i.e., someone looking for the best clinic probably does not care that the best one is miles upon miles away.)
Forum discussion at Local Search Forums.
