Google Traffic Source Distribution

Barry Schwartz at the SEW blog points to an interesting Hitwise breakdown of Google traffic by property:

//weblogs.hitwise.com/bill-tancer/Google%20Properties%20Table.PNG” cannot be displayed, because it contains errors.

source: Hitwise

Google Maps (the "new" Google Local) is #5 and Google Local (the "old" Google Local) is #16 in terms of where the traffic and usage comes from. Combined they represent 0.87% of Google traffic. Here's a comparison of the two domains from Google Trends:

google maps google local

Again, this isn't an absolute representation of traffic, it's a reflection of the number of people typing in these queries but it validates the Hitwise data above and argues that Google made the right choice (at least in terms of consumer awareness) regarding what to call its local product (Maps).

The Hitwise data above also points to something else that's interesting in my mind. The marketplace is sort of compelling Google to "verticalize" in order to better satisfy user queries and provide a richer, more structured search experience. (There are also clear advertiser benefits to this.) Yet building out separate vertical domain after separate vertical domain apparently isn't "working" for Google — in the sense of direct navigation and adoption. That's why we may have Co-op, the all-purpose vertical generator, vs. separate sites for each conceivable vertical: e.g., Google Autos, Google Jobs, Google Houses, etc. Indeed, Base and Co-op are working the same "problem" for Google from two different angles (thanks Marty Himmelstein for pointing that out to me in an email.)

In the context of where Google's traffic is (Google.com) it makes sense for the company not to create separate labor-intensive vertical destinations but instead to try and deliver richer vertical environments off the main SERP (there's a whole discussion here about whether Google can/will be able to effectively do this with Co-op and Base). This “one box” approach isn’t incompatible with separate vertical destinations; for example, shopping query results tease Froogle, news-related queries that tease News, local queries tease Maps, finance-related queries tease Finance and so on. But it would appear that going forward Google is going to keep more users within the “Google.com” environment and build more vertical experiences directly into that.

It’s probably also the case that Google’s “ancillary products” don’t have tons of adoption because the company doesn’t aggressively promote them. Yet another way to look at the Hitwise traffic distribution table above is: “The Head” vs. “The Tail.”

____

Related: Om Malik's somewhat cynical take on the same data ("Google Reality Check") 

One Response to “Google Traffic Source Distribution”

  1. Whazzup with Shopping at Google? « Screenwerk Says:

    [...] What Google appears to be universally doing (and for some time now) is disambiguating queries and threading “vertical” content into search results (local, products, news, movies, etc.) followed by a more specialized environment built around the content type. Co-op also plays into this “verticalization” strategy. [...]

Leave a Reply