The mobile Web is a dynamic and challenging environment. One of its biggest challenges is that conventional Web analytics applications don't do a good job of tracking mobile Web site usage. That makes it difficult to tune and enhance your mobile Web site's performance. Today, we'll talk about some of the different mobile analytics methods, and the strengths and weaknesses of each approach.
Growth in the Mobile Web
Mobile Web usage continues to grow like a weed, in spite of its challenges. For example, according to comScore 30 million U.S. users accessed the Web using a mobile device during the month of January 2007. In that same month, 159 million U.S. users used a PC to access the Web.
So from a reach perspective, the mobile Web is already about 19 percent of the size of the PC Web market. Of course, users may not use the mobile Web as often as they use the traditional Web, but this is still significant.
In addition, an ABI Research Study projects growth in the mobile browser market from 76 million units in 2007 to almost 700 million in 2013. In addition, browsing capabilities will expand significantly. The bottom line? The mobile Web is going to be a hugely important environment.
Mobile Web Analytics
With such a large and growing market, it's reasonable to presume that more people will begin to offer Web sites targeted at the mobile Web. As this takes place, and as usage begins to grow, demand for quality analytics data will increase.
Log file analysis is another approach that can be used, but there are limitations with that as well. One limitation is the inability to collect mobile-specific information, such as the handset capabilities. Another problem: all visitors with the same carrier and phone type will be indistinguishable from one another.
But, there are some other options out there. Here are some of the approaches from mobile analytics leaders:
- Packet sniffing. This approach relies on installing an additional server (the "mobile analytics server") into your Web environment. A switch is then used to pass a copy of the incoming packet data to the mobile analytics data, while allowing the same packets to pass through to the mobile Web server without any delay. In addition to basic Web data, you can get information on handset resolution, the mobile operator, handset type, and browser type. Australian based Amethon uses this approach.
- Site redirection. This method relies on traffic being redirected through a different server, and then redirected back to the mobile Web server. This also allows for the collection of a rich set of analytics data, similar to what you get with packet sniffing. This approach is used by Bango.
Web analytics is notorious for having accuracy problems. As Greg Harris, CEO of Mobile Visions, tells us, the situation with mobile analytics accuracy is no different.
Accuracy, however, isn't the main issue. While it would be great to get fundamentally more accurate data, it's not available. As with traditional Web analytics, you get far more data than those used by traditional marketers to track TV, radio, and print campaigns.
This data can be used to help mobile Web site developers understand how their content is being consumed, detect problems with their sites, and find out how to improve conversion.