Hulubaloo over Nielsen figures
Online video site Hulu has questioned the accuracy of Internet traffic figures produced by Nielsen Media, one of the major audience measurement firms. Hulu says numbers which are incorrectly low threaten its ability to get a fair share of advertising spending.
The complaints spark from the figures for April for which Nielsen claims Hulu had 8.9 million unique visitors. (That’s an attempt to count how many different people visited a site during the month.) However, another firm, comScore, put the figure at 42 million.
Hulu is also questioning the fact that Nielsen shows its unique visitors numbers falling consistently between February and April (9.5 million > 8.9 million > 7.4 million) but that the number of video streams – that is, clips watched – has risen in the same period (309 million > 348 million to 373 million).
That apparent contradiction could be explained by individual users watching more clips, for example if the site has a better range of content. However, that would mean a pretty significant change in behavior with the average viewer watching over 50 percent more clips per month.
Nielsen does not measure actual traffic in its entirety. Instead it operates a panel system, similar to that for producing TV ratings, in which it monitors the activities of a sample group and then extrapolates the results. (As the participants are aware they are being monitored, the Nielsen figures are probably particularly understated for adult Web sites.)
One of the big questions in online traffic is whether such a system, which works well for a few broadcast TV channels, and passably for a few hundred cable channels, is still adequate on the internet where you are comparing millions of Web sites.
The ReadWriteWeb site notes that other measurement firms have differing figures, with Quantcast putting Hulu at 26 million in April. Quantcast’s system is arguably more accurate as it involves putting code on a site and directly measuring all visitors. The downside is that this means it can only put together figures for participating sites, while systems such as Nielsen can cover the entire web.
Of course, the real key to online traffic measurement is not so much the raw numbers but the trends. What’s really important is that the figures accurately portray how rival Web sites perform in comparison to one another, and whether a particular site’s audience is moving up or down. If Nielsen’s figures are shown to be failing in those tasks, it would be a major blow to its reputation.
Related Posts:
May 15th, 2009
This is a fascinating example of how a traditional media company, Nielsen, has been very, very slow to understand and embrace–let alone leverage–the Internet.
Nielsen’s rating algorithms have been subject to skepticism for ages, but it is clear that they don’t really know what they are doing in the online arena. The NYT story on this today reported a difference between 9 million visitors reported by Nielsen versus 42 reported by another organization. I myself watch almost all my TV online since I can watch my shows on my own schedule and thereby also bypass Tivo or some other expensive gadget. I can hook my laptop (with Blu-Ray drive, to boot) up to the HDTV to do so, too, if I want a big screen experience. I’m not one of a mere 9 million, I can assure Nielsen of that.