Mark Ritson: Dodgy video metrics prove TV is way ahead

Why don’t purveyors of online video measure average audience per minute? Because it would reveal they are thousands of times less popular than TV.


The tsunami of bullshit that is digital video continues to head inland, threatening to subsume TV advertising and much of the marketing hinterland in 2017. The bullshit comes not as one single giant flood, but in a series of intermittent waves. Each containing slightly more sewage than the last. Each arriving without challenge or counter from the marketing community. Each biasing marketers against so-called traditional media and in favour of new, hot ‘digital’ alternatives.

Last week it was the coverage of the American presidential election and the promotion of BuzzFeed’s live election night Twitter broadcast. It garnered the now standard fawning headlines from the usual suspects. Adweek noted that “BuzzFeed Didn’t Need TV to Make a Mighty Fine Election Night Show for Millennials”, Reuters claimed “Twitter impresses advertisers with BuzzFeed US election livestream” and Variety drew attention to the “6.8 million total viewers” drawn to the coverage.

BuzzFeed actually went even further and claimed that its broadcast was the “most viewed US election-related live stream on Twitter”, which is somewhat akin to me claiming to be the best looking Cumbrian marketing professor born in 1970 (whose middle name is Blaylock).

READ MORE: Mark Ritson – Facebook’s erroneous video metrics show no one has a clue about digital

At first sight 6.8 million viewers certainly appears an impressive result. But as usual with digital video it’s only impressive if you use ridiculously self-serving horseshit metrics. All we can infer from the 6.8 million is that at some point last Tuesday night for at least three very brief, probably partial, and likely soundless seconds these people chanced upon the coverage on Twitter. When we turn this into a proper measure of viewers, i.e. an average audience per minute, the 6.8 million becomes 165,000.

In American media terms that number equates to what statisticians often refer to as “almost fuck all”. According to Nielsen, 71 million Americans watched the prime.time coverage of election night last week meaning Buzzfeed and Twitter managed to snare a grand total of 0.2% of the viewing population. CNN won the night with 13.3 million viewers, meaning that for every lonely viewer watching the election on Twitter, 80 were watching it on CNN. That’s not to mention the additional 73 watching Fox News or the 67 tuning in on NBC.

6.8 million viewers certainly appears impressive. But as usual with digital video it’s only impressive if you use ridiculously self-serving horseshit metrics.

And for those of you totally addled digital marketers who are now thinking “ah yes but millennials don’t watch TV anymore, they need Twitter”, well you are kind of right: 83% of those watching the election coverage on Twitter were indeed under 35. But that equates to only about 137,000 millennials in total. Given CNN managed to attract 3.2 million millennials to its coverage, that puts that whole argument to bed.

By now, of course, we are used to the digital video shenanigans that use incomparable metrics to perpetuate the myth of growing audiences for digital video and declining ones for TV. We saw it during the World Cup when ESPN claimed an average audience of 4.6 million for its TV coverage and 115 million views for its digital coverage of the tournament. But the 115 million became 300,000 when Nielsen turned it into an average audience per minute.

READ MORE: YouTube takes a swipe at Facebook over silent videos

We saw it during the Olympics when NBC boasted 3.3 billion streams of its digital coverage of the Olympics and tried not to admit this equated to only about 3% of the audience when it became a proper per-minute measure. And we are seeing it with American football, where Twitter’s much heralded broadcasts of NFL games are achieving average audiences of 250,000 compared to the 15 million sad losers who did not get the memo and continue to watch the games on their TV.

Of course those last two paragraphs do something that most media coverage assiduously avoids: they present the TV and digital data side-by-side using the same metric of average audience per minute. Why we continue to obfuscate video audience measures with an apples to oranges approach is beyond me. Can I suggest that if Twitter was getting 50% or 75% of the numbers achieved by TV, rather than a tiny rounding error, they would be promoting their average audiences per minute in flashing 8 foot high inflatable lettering on top of their HQ.

There is something interesting about the reluctance of audiences to get with the programme, leave TV and embrace digital video. Could it be that digital video will struggle to offer the big screen glamour, content expertise and social context that has made TV the dominant medium for the past half century? I smell a real battle in the wind for 2017. One that, for once, Google and Facebook may not be well placed to win. If you can’t beat ‘em buy ‘em may well be the conclusion of the digital duopoly by the end of next year.

As they used to say in the TV game, watch this space.