Wednesday, March 4, 2009

Views of Political & News Videos Up 600% -- Now What?

YouTube's news manager told BeetTV yesterday that views for news and political videos were up 600% this year. That's good news for journalism. It means that people are interested in news and politics. Or at least can be engaged under certain circumstances. And if news organizations can leverage that insight, they can increase their views-- and, hopefully, their revenue.

This is rich information we can use to help shape the future forms of journalism.

But first, here's how not to use this information:

Editor: "Political and news videos are up 600% on YouTube! Reporter people: Make me more videos about what's going on at City Hall!"
That's a really good way to waste your valuable staff time on stuff that'll never get seen.

It's like a food manufacturer learning that sales of Duncan Hines' cake mixes are up 600% and immediately ordering its staff to start producing more cake mixes. All kinds (chocolate, vanilla, red velvet...). Year-round. And then ending up perplexed about why their mixes didn't sell. The problem is, if the manufacturer had done a little digging, they'd have found that the majority of that 600% was for sales of chocolate cake. At Valentine's Day. See what I'm getting at? You have to look under the hood of the data to understand exactly what's driving the increase and where the opportunities lie. Only then can you create an effective video strategy for your news organization.

So here are some of the questions I'd ask about the 600% increase:
  • What types of videos were people watching, and what does the pie chart of types look like? Ie: What proportion of views was for:
    -- Campaign-created candidate-speaking-to-the-camera videos?
    -- Campaign-created ads?
    -- User-created artistic creations (songs, parodies, etc...)?
    -- Professionally-created artistic creations (SNL, etc...)?
    -- Raw footage from professional journalists?
    -- Official reports from professional journalists?
    -- Debates shown end-to-end?
    -- Snippets of debates?
    -- Etc...

    This tells you what kind of videos are worth producing (or enable to be produced and posted, if user-generated ones are among the most valuable).

  • What proportion of the navigation paths started outside YouTube vs. started inside?

    Ie: What proportion of video views:
    -- Started with someone arriving from an outside link?
    -- Started with someone doing a search on YouTube?
    -- Started with someone clicking on a video in the "Related Videos" list?
    -- Started in YouTube channels? (And what particular channel drove large numbers of viewers?)
    -- Happened when someone decided to re-view a video?

    If a significant proportion of video views started with someone using the "Related Videos" list, then you know that, to maximize the number of video views on your site, you're going to have to make sure your interface does a good job of drawing visitors in. Ie: The editor might get more bang for his buck by asking to the web team to beef up the site's interface, rather than simply asking the reporters to create more videos.

  • What kinds of videos did people watch all the way through, vs. what kinds did people leave in the middle?

    If you assume that the videos that people watched all the way through were the ones they liked (and the ones they didn't were the ones they didn't)*, then you'd want to analyze the videos to figure out what it was about the first set that people found interesting and the second that they didn't. And then use that information to shape your own videos. In order to create happy "customers" who are likely both to return and to recommend your videos to other potential viewers.
It is exciting that views of news and political items went soaring last year. Now let's use that information intelligently to help news organizations craft effective strategies.

* And you'd have to test even that assumption...

No comments: