One of the ways that Names@Work distinguishes itself from search-engine-only or blog-only marketing companies is our knowledge of domain names. I’ve been very interested to find data about domain names as a driver of traffic.

Today, in Paul Sloan’s article Masters of Their Domains (are we ever going to get an original title for an article on domain names?) quotes an anonymous “Yahoo official” as estimating that 15% of all traffic on their network comes from “type-ins”, or domain names. For reasons I don’t understand, neither Google nor Yahoo will release their official tallies.

Super unofficial talks with people at VeriSign netted us a super unofficial figure of somewhere between 10 - 20%, so that fits. DN forum suggested a conservative 6%.

Another way to look at the problem comes from a study at UCLA (undated, but citing studies done in 2004) called Analysis of User Web Traffic with a Focus of Web Activities. This study, which took pains to eliminate factors (such as web crawlers) that don’t measure use by actual people, found that 13.6% of all web traffic is under the direct or indirect influence of search engines, and that 6.8% of the time users start their web sessions using a search engine. In other words, 86.4% reached web sites by non-search means. Hmmm.

The study found that in one sitting (average 2 hours per session, 5 minutes per web page), a typical user visits about 22 web pages on 5 different web sites. Of great interest to our investigation, they found that during this sitting, the typical user went to a web page directly (either by typing in a URL or by using a bookmark) 3 times during the session.

Of the 22 pages visited during a typical user session, 3 of them, or 13%, are found by direct traffic. The study doesn’t examine what percentage of the 5 web sites visited during a session are from type-ins and bookmarks, but I will boldly assert that most people type in fully qualified domain names, which means new web sites — so that of the 5 web sites visited during a session, a substantial portion come from direct traffic.

Conclusion — the “unofficial” 15% of all web traffic coming from domain names or bookmarks sounds about right.

Questions to examine:

1. How to distinguish between bookmarks and “type-ins”?
2. How many people type in a domain name into the search engine search box? I suspect many people don’t understand the difference between the URL box and the search engine box. Traffic counted as search engine traffic might in fact be domain name traffic typed into the wrong box….

Posted by Antony in: Domains, General, Search