The struggling wireless Web has a new worry: Nobody really knows how many people are using it.
Whether for bragging rights, ways to entice advertisers, ways to attract the best content, or a combination of all three, wireless industry players want to know who is using the budding medium.
One way to find this out is to use traffic reports, through which the number of visitors to a Web site is measured. Although these reports are a regular part of the wired Internet, the same type of research is proving difficult, and unreliable, when it comes to the wireless Web.
Compounding the problem are the wireless carriers themselves. U.S. carriers like AT&T Wireless monitor the traffic to their own wireless Web pages to gauge the types of games or news stories that people want. But none of the carriers makes public just what these figures are, all citing competitive reasons.
What ends up missing is a third-party perspective, and companies that want to boast about their success are quick to take advantage of that.
For example, Yahoo Mobile issued a statement recently claiming to be the No. 1 mobile site in the month of May, outpacing the nearest competitor by more than 150,000 clicks. The company made its claim based on how many people visited mobile.yahoo.com. This site is a traditional wired Internet site that lists Yahoo wireless services and other mobile Web sites to bookmark, as well as information on how to buy an Internet-ready phone.
A Yahoo spokesman said the company was trying to gauge wireless traffic from the number of registered users at the site. While others offering the same type of mobile portal, including Microsoft and AOL Time Warner, wouldn't comment on the findings, some in the industry were skeptical that that figure crowns Yahoo as the king of mobile.
"Can Yahoo claim the prize? Only if they are brave enough," said Alan Jones, Webmaster for AllOutWap.com, which ranks mobile Web sites based on traffic figures culled from wireless server data. But even Jones concedes his company's rankings are inaccurate, warning those who use the information that the numbers are only estimates.
"Using the current crop of phones and given the levels of technology, it is absolutely impossible to prove, or in fact disprove, (Yahoo's) claim," he said.
Some of the bigger players in the traditional Web traffic measuring industry deliberately are not venturing into the wireless arena, due to technology obstacles and low wireless Web usage thus far.
Nielsen/NetRatings and Jupiter Research both say it's technologically possible to measure wireless Web traffic. One solution is to do what both companies have been doing for the wired Web: Pour the tracking software that would normally go on a personal computer into a cell phone.
So far, no handset manufacturer has agreed to do so, although a spokesman for Nielsen/NetRatings said several have been approached.
Rebecca Young, marketing vice president at Nielsen/NetRatings, said another reason why the company isn't diving into wireless tracking is because the wireless Web has been slow to catch on. For example, KDDI, a Japanese wireless carrier, recently said it saw a decrease in new wireless Internet subscribers, a drop of about 30 percent in June compared with May figures.
"We are staying engaged in the discussions," Young said. "But in no terms are we pursuing anything. What it really requires (is that the wireless and tracking) technologies be ready and capable."
A spokesman for Jupiter Media Metrix said the company isn't ready to jump into wireless measurement just yet either.
If the traffic-monitoring companies ever do begin measuring wireless Web traffic, they are going to have to solve a sizable riddle. Using current methods, AllOutWap's Jones said, estimates of just how many people go to a wireless Web page can be off, sometimes dramatically.
The problem lies with just how a cell phone actually views a wireless Web page. When a mobile Web user actually accesses a Web page, the information is usually stored on a server. This way, the phone's memory isn't taken up with the Web page information.
While that makes surfing easier, the counting of traffic suffers. Jones explained that a mobile Web page accessed by one person is sometimes cached for up to four days at a time. Hundreds of people could access that page in those four days. So hundreds or even thousands of hits are never registered.
There are ways around this, Jones said, such as putting information on the page that expires in a short period of time. Those methods have proven unreliable, however.
"But in practice, you can't rely on the systems doing what you want them to do," he said.