New Research Reveals More Flaws in Web Tracking
Rich Gordon offers further proof that Web traffic measurements are deeply flawed. In a post on Poynter's E-Media Tidbits, Gordon points to recent data from comScore:"Almost a third of Internet users delete 'first-party' cookies from their computers at least once a month. Also, seven percent of users delete these cookies four or more times per month, comScore found."As Gordon notes, things get dicey when you consider that many analytics programs use cookies as a measurement tool for unique visitors; a key metric for advertising rates and overall estimations of Web site popularity. If people are deleting cookies and analytics programs are tracking cookies, then it's entirely possible that unique visitor data for many sites is inflated (i.e. users are counted more than once because a new cookie replaces the deleted one).
This cookie conundrum is only one component of a Web-wide tracking issue that extends from big publishers to small-fry bloggers (like myself). If publishers can't trust the numbers and advertisers can't trust the numbers, how do we set ad rates? How do both parties -- publishers and advertisers -- get a fair shake?
To me, the answer lies in the "Average Time Spent" metric. Granted, this metric can also be tweaked and miscounted (especially if it's aggregated via cookies), but my simple-minded nature continues to latch on to one key truth: the longer people stay on a site, the better the chance an ad/message will be seen (unless it's a punch-the-monkey ad -- no one should see those).
Even if cookie usage can be trusted, established metrics like page views and unique visitors don't benefit targeted sites (like blogs). To illustrate: let's say I run a "boutique" site that caters to a niche audience. The site gets 200,000-300,000 page views a month and attracts around 50,000 visitors. These figures aren't particularly enticing to most advertisers, even ones in my specific niche (assuming there are any). But my "Average Time Spent" figure is an ace in the hole: it hovers in the 2-3 minute range, which is significantly higher than the mere seconds most folks spend on Web pages. Shouldn't this metric count for something? If I can show advertisers that my targeted audience is spending legitimate time on my site, shouldn't that somehow translate into a reasonable ad rate?
Under the current system, folks with page views in the 200-300k range don't get a second look from advertisers or agencies, yet these same sites may be attracting the exact readers certain advertisers seek.
As I see it, there are two core problems that need to be addressed here:
- Technology -- We need a simple, reliable analytics system that -- more times than not -- gives us numbers publishers and advertisers can trust.
- Metrics -- We need to move beyond the page view and the unique user and see the value in other metrics. We also need to apply the correct metric for a particular site/audience/advertiser. What works for a big publisher might not work for a smaller one, and both sides -- publishers and advertisers -- need to understand this. (Note: Recent news suggests that change may be afoot).
Disclosure My own sites fall into the "boutique" category I discussed and I fully acknowledge that a move toward "Average Time Spent" would boost my own revenue. Nonetheless, I believe this is a conversation that needs to take place and I'm pretty sure I'm not the only one who feels that way.
More Recent Stories:
• iPods and Violent Crime, Together at Last
• The Real Format War Has Yet to Begin
• New Currency in a Free World
• Web Analytics Stay Ridiculous
• Time for the "Oh My God It's Cold" Stories
• Memo to Jerry Yang: Use Your Shift Key
• Tutorial: Create a PDF with a Web Browser
• What's Behind that Multimedia Presentation?
• Microsoft Courts Yahoo with $44.6 Billion Deal
• Wi-Fi Comes to Boston Commuter Rail
