Today I am publishing the transcript of my recent interview with Carter Maslan. This post summarizes some of the main points of the discussion. We spoke quite a bit about the new service area business tool from Google. Some of the major points made by Carter about this were:
- Many businesses don’t want to have their address listed in local search results. For example, a plumber who works out of his home, but always goes to the customer to provide his services. Carter also noted that service area businesses are “primarily defined by whether or not the business brings its services to the customers”. However, service areas businesses can include consultants that work from home that the customer can call to obtain their services.
- The ability to set oneself up as a service area business is “pretty broadly accessible”.
- “Even though we don’t have specific numbers to share … there area ton of home-based and service businesses in this country”, and “I think it is at least a third of the overall total”.
- “giving the end user a PO Box as a pin on the map is not really helpful”, and “If a business really cares about its customers knowing where its PO Box is, I think it’ll be clear that it should be a service area business”.
- I also asked Carter about spam. He indicated that they do more or less the same thing they do with other types of spam. For example, when I asked him about a plumber who declared they would serve anyone within 1,000 miles, he indicated that “There are a lot of signals regarding whether or not this is suitable”.
We also discussed Place Pages. What emerged from the discussion was that the purpose of this was to provide access to all the available information about a given place. This would include, but not be limited to, businesses that have no web site. The information on a Place Page could include information provided by the business, but will also include information found by Google in crawling the web.
I asked Carter whether or not having individual landing pages for each location of a business with many locations was preferred. The answer was yes, provided that there was meaningful information that differentiated on locaiton page from another (inventory info, driving directions, etc.)
Google has also made it quite a bit easier for people to report errors. This is basic crowdsourcing in action. They are happy to take reports even if all they specify is that something is wrong with a listing. Note that Google can also look at user interaction data (with a particular listing) to get signals to possible problems as well.
We talked about training the local search algorithm. One concept we bandied about was that of having humans build a mini-data set (e.g. some number of tens of thousands of hand researched listings), and then running the algorithm to see how its results compared to the hand crafted test set of data.
As always, in the interest of providing a short synopsis, I have passed over many details and other points from the interview. Read the full interview for more.