The long form

, Friday, March 21st, 2008

Since the advent length of YouTube, we’ve been advising people that the average user attention span is about 4 minutes, and that making anything longer just means that people will click away.

But this political season has been challenging that assumption, when some of the most viewed videos have been extended speechs of thirty minutes or more.

Are we reaching the turning point for web video where users are willing to watch longer content? Are people willing to reach the Illiad in addition to a haiku? And how do you test the proposition of a long form piece of content versus a shorter one.

I explore these questions below the fold…

Some of the most popular videos on YouTube have been the speeches given by the various presidential candidates. Obama's greatest hits Unsurprisingly, Barack Obama’s campaign has been the foremost here – both his mastery of the spoken word and leadership online combine to explain why his speech from Tuesday has seen over 2.5 million views on one version alone. Previous speeches have also garnered significant attention: his 14 minute victory speech after Iowa hit 1.1 million views, his 13 minute speech after New Hampshire scored 800,000+ views, his 34-minute speech on the day before Martin Luther King day garnered 725,000+ views, and his SC victory speech was viewed just under 570,000 times. And those are just the most popular versions of each in the official channel; with 800+ videos, I’d be interested to see the total number of net views for videos longer than 10 minutes and those shorter.

While McCain and Clinton also have videos that are over 10 minutes long, they’re significantly less popular – Clinton’s most popular long video (26th overall) is a 17 minute interview with letterman which has been viewed 111,000 times; McCain’s most popular long video (4th overall) is a 12 minute ad which clocked around 110,000 views. This might just be a function of not promoting their speeches as much as other video, though their top videos were also much less watched than Obama’s top video.

Remember also that YouTube is particularly unforgiving when it comes to counting views. YouTube only counts those viewers who watch a video all the way to the end as a view, so anyone who bails out early (which is more likely with a longer video than a short one) won’t be counted.

Why is it, then, that these videos defy the conventional wisdom on YouTube? Maybe a part of that is the media’s obsession with sound byte politics (speaking of which, the Facebook Politics app which I previously panned for its ABC content is reaffirming my criticims by asking users what they thought of the Rezko thing with no mention of any of the candidates recent policy statements). Given that newscasts feature only snippits of speeches, voters might want more than just the readers’ digest version of what a candidate’s about, so they turn to YouTube. This is yet another example of how users have revolted against letting the media control how and what they watch, like TiVo and BitTorrent.

So should your non-profit think rating about moving to longer videos and abandoning the 4-minute length guideline? It’s tough to say. One thing that differentiates these longer videos by candidates is that it’s the candidate him- or herself speaking. Most non-profits don’t have a single monolithic face who can be featured for that length of time.

In addition, YouTube is a difficult medium to measure effectiveness in. While a video might shoot up to the top in number of views, it could be because its embedded by someone who wants to attack it – so view count is hardly a measure of success. Rather, I’d argue that counting the number of friendly embeds is a better metric. For users to watch and not only want to pass it along, but to be committed enough to feature it on their blog or website means that they’ll be your evangelist in a committed way.

Testing this is a tricky question – splitting the list in half isn’t really viable with so small a sample size. Instead, consider sending both a long and short version of a given message to supporters, and ask them to embed whichever they prefer. The catch here is that you’ll be testing what will, in theory, be the best content from a video against the rest. But it should at least give you an idea of what percentage of your supporters would post a long video for you.

I think it’s still best, at the moment, to assume that short videos will do okay. But I no longer believe that they’re the only way to go with YouTube.

Comments are closed.