Google Maps VP Explains New Nearby Places Feature
Mike Blumenthal did a rather nice Q&A with Carter Maslan, Google's vice president of product development for Google Maps, about Nearby Places You Might Like, a new feature that surfaces similar stores when you search for a business on Google Maps.
I wrote about my issues with this feature Feb. 7, noting that it showed too many competing businesses on the Place Pages.
I still believe this to be true, and wish to see more "related" businesses. So if I search for Starbucks, I won't just see competing coffee shops, but other places to get food. Sometimes, this works well. Take this search for Starbucks, in Trumbull, CT:
Instead of all coffee shops, I get a Subway listing and other eateries in addition to Dunkin' Donuts and other coffee shops. This is much more useful than a list of nine or 10 java joints.
In Maslan's own words, Nearby Places is geared to help you "find and discover places you'd like to know about. The feature is designed for everyone searching for places, whether they be stores, transit stops or historic landmarks."
Maslan said Google is exploring the "relatedness" among different places, which is the key to mapping out relationships among the different businesses in cities and towns.
What's clear from the interview is that Nearby Places is rather incomplete. As Blumenthal noted, some places you search for on Google Maps don't render any Nearby Places.
Maslan said the Google Maps team is still tweaking the feature, noting that Google hasn't collected enough info for all of the queries readers might make.
Blumenthal asked a lot of questions, some of which got similar stock answers, but it's interesting to note that Google is attacking Nearby Places You Might Like as a search algorithm challenge that will be constantly revised and improved.
I think we forget that, even though Maps are visual representations, Google is increasingly aggregating a lot of traditional text search data in the results. As Google improves its algorithms, better results will follow.