According to BARB, 12.09m people watched Eastenders on Friday 1 January. How does BARB work this out?
“Viewing estimates are obtained from a panel of television owning private homes representing the viewing behaviour of the 26 million TV households within the UK. The panel is selected to be representative of each ITV and BBC region, with pre-determined sample sizes. Each home represents, on average, about 5,000 of the UK population.” (Source: http://www.barb.co.uk/about/tvMeasurement?_s=4)
OK. Now imagine we worked out web figures on such a low tech ‘sample’. Actually web stats are surprisingly accurate — products such as Omniture and Google Analytics track every page impression pretty accurately, which means that to a very small percentage, we know exactly how many pages were displayed across a website.
Marketing folks are often surprised how low some web ‘traffic’ (in TV speak, read ‘viewing’) figures are. But actually they’re very accurate, especially compared to the totally inaccurate method of TV stats.
The next generation of web applications are already posing further problems. Two years ago we pitched to a very well known football club to look into redeveloping their website and other Internet applications, once of which was SecondLife. They specifically asked us to recreate their stadium in SecondLife. During the pitch, our Creative Director boldly stood up (as Creative Directors do) and showed a screenshot on SecondLife of their stadium, already in the virtual world. His point was twofold – 1. there was already an unofficial one which the club could ‘buy’ from the obviously-bored fan; and 2. he was the only person in the stadium, which had about 45,000 ‘seats’.
- TV stats are inaccurate
- Web stats are accurate at the expense of appearing lower
- Finally, when brands or Rights Holder start looking at ‘concurrent’ stats, they’re actually much lower than even the web stats.