PDA

View Full Version : comScore to take Web 2.0 seriously



grchl3
12-26-2006, 10:28 AM
http://img413.imageshack.us/img413/4606/keyboardvc2.jpgTelevision has Nielsen, but the web has no clear leader for ranking the world's busiest websites. There's no shortage of competitors in the space, of course, and you've probably heard of many of them: Alexa, comScore, HitWise, and the newcomer, Compete. None of them provide uncontested results, however. Many of them do worship the almighty "page views" god, but that's about to change.

In the wake of announcing that MySpace has taken the top spot in page views online from Yahoo's clutches, comScore has said that they are working on new metrics that will also take into account the trappings of Web 2.0, including interactive AJAX-driven web pages which do not necessarily generate page views. The problem is simple: the number of page views on a site does not indicate how much a site is used (let alone valued) by its readers. Google Maps is a fine example: after initially visiting the site, actual page loads are few and far between, as the application pulls data dynamically from Google as needed. Yahoo, too, has been using AJAX and other interactive approaches to content, which some have argued invalidates MySpace's climb to the top.

In a statement issued last week, comScore acknowledged the problem, saying: "The recent decline reported by comScore in Yahoo! Sites page views underscores another emerging issue in the Web metrics measurement industry. New technologies such as AJAX—which enable real-time site updates without needing to refresh a page—are impacting the relevance of page views as an accurate measure of the intensity of consumers’ Internet usage. Yahoo! in particular has begun implementing AJAX and other Web 2.0 technologies across their sites."

Dr. Magid Abraham, President and CEO of comScore Networks, added: "While page views will not altogether cease to be a relevant measure of a site's value, it's clear that there is an increasing need to consider page views alongside newer, more relevant measures. comScore is proud to continue carrying the torch as an industry innovator with the development of a new suite of metrics that will effectively address the Web 2.0 landscape by including enhanced measures of user engagement and advertising exposure. We will be introducing these new metrics to the industry in 2007."

Just what those changes are remains to be seen, but we're eagerly anticipating them. Why?

More than bragging rights

For better or for worse, web ratings are needed. They help good sites stay afloat, but the magical nature of these metrics leaves many site operators, advertisers, and increasingly, bloggers, scratching their heads.

If comScore says you have 1 million readers and 3 million page views, but you know that you have 2 million readers and plenty of interactivity not reflected by page views, it can cost your website money. Advertisers might look down their nose at you. Companies may chose to ignore you. Generally speaking, you're not getting credit for your traffic. This hurts some more than others.

In fact, I know of more than one major acquisition deal involving hot Web 2.0 sites that have been stalled on this account. Site owners aren't about to leave money on the table because of what an analytics firm says, but companies in acquisition mode aren't always looking beyond the metrics. What's worse, the inaccurate numbers give a false impression of what's really attracting and retaining users online, and it's particularly unfair to the most cutting edge sites. Consider MySpace: do they really deserve the #1 spot simply because to do anything on MySpace, you have to load many, many pages? MySpace was built to generate page views. Other sites are built to minimize them.

To make matters worse, many of these traffic measurement services use data collection methods which are known to undervalue certain classes of users, including those dedicated to technology. comScore, for instance, tracks users who voluntarily participate in an e-commerce tracking system, and then it uses behavioral data from those opt-in users to extrapolate trends on a massive scale. As you might guess, this means that some sites are under represented. Ars Technica, for instance, attracts highly savvy readers who by and large do not opt-in to such systems. The same is true for most technology sites online, and their scores suffer; comScore does not truly know our audience, only a small portion of it. Other metrics might rely on a toolbar or similar opt-in systems, which immediately raises the question of what qualifies as a "standard" Internet user. If that definition remains "guy who buys things online using Internet Explorer with this special toolbar installed," then you can see the problem.

With comScore showing interest in taking Web 2.0 more seriously, perhaps there's hope that someone will eventually crack the web-ratings nut.

:source: Source: http://arstechnica.com/news.ars/post/20061225-8493.html