Flawed Traffic Stats Impede Hulu's Growth

By Mike Schuster Jul 27, 2010 12:50 pm

Hulu and online video as a whole could suffer because of incorrect viewer measurement.

For any SEO manager, wildly fluctuating traffic numbers are just as bad as a verified drop. Reliability is paramount when sizing up a website's audience and can be the difference between signing an advertiser and never hearing from them again. But even after nearly two decades of recording page counts and unique hits, the stats are still flawed, leaving traffic numbers to artificially rise and dip month to month.

And no one is more aware of the risk of an artificial drop than Hulu.

Hulu has enjoyed a lengthy stay as one of the most popular sites in the US, thanks to ComScore's (SCOR) measurement service. But ComScore -- an industry leader in the field of Web traffic -- recently overhauled its methods of pulling visitor data. And as a result, the numbers for Hulu -- an industry leader in the field of online video -- dipped precipitously from May to June of this year, according to the Los Angeles Times. Hulu's numbers in May hit an estimated 43.5 million while, post-overhaul, ComScore recorded a mere 24 million in June -- nearly a 45% drop.

Since it's unlikely Hulu's traffic took such a free fall over the course of June, the fault lies with the data gathering. And when visitor stats are gathered from several sources, the numbers can be essentially inadmissible.

Fellow video-streaming site Justin.TV was subjected to such ambivalent data when viewing its global audience between ComScore and competing tracker Quantcast. According to ComScore, Justin.TV enjoyed a healthy growth from 12 million to 15.6 million over a March to June period. However, Quantcast recorded a severe plunge in traffic -- 26 million to 17.2 million -- over the same period.

"We have a spreadsheet internally that we keep updated with third-party numbers. I am as surprised as the next guy when we update it," Evan Solomon, Justin.TV's vice president of marketing, told the Times. "It's like opening a box when you don't know what's in it."

And like Justin.TV, Hulu depends on these numbers to pull in advertising dollars that it needs to not only survive as a site, but prove the potential and profitability of online commercial video. Such is the burden of one of the most popular video-streaming sites around.

Unlike radio, TV, and newspapers, Internet traffic is gathered from a variety of sources, and the fluctuating data leaves much of it open to interpretation. That still doesn't sit well with Hulu's backers -- NBC (GE), ABC (DIS), and FOX (NWS) -- and the advertisers hesitant to quantify dollars against online viewership. Hulu's paid subscription service Hulu Plus is still in its infancy, so the free, ad-supported model remains the site's chief source of revenue.

Considering the ongoing squabbles between content producers over Hulu's ad revenue -- resulting in canceled plans for a UK launch, Viacom's (VIA) removal of Comedy Central content, and the birth of Hulu Plus -- accurate traffic numbers could be the tipping point for studios and advertisers to nail down a fair deal. But with the rising popularity of Netflix's (NFLX) streaming content, the ubiquity of Apple's (AAPL) iTunes video, and Google (GOOG) TV looming on the horizon, accuracy needs to happen fast.

Then again, is that even possible?

Minyanville's general manager of online marketing and development Annie Stickney said, "Every data collection source is always on average 15% different from another one, according to the laws of standard deviation."

But regarding ComScore specifically, she has a reason to doubt the team's numbers. "They mentioned that they have not figured out how to accurately track people who use Macs," Stickney said. "If anyone these days produced a site that was inaccessible to Mac users, that website would no longer exist. So why is it fine for them market an analytics and user tracking program that does the same?"

Sounds like Hulu should be asking the same question.