Real-time curation vs editorial judgment, or, the redundancy of humans
Michael Arrington at TechCrunch famously suggested that if one were to build a news organization from scratch, one path would be to hire away the top 50 writers at the New York Times without all of the attendant cost (pensions, printing plants, editors, etc) and put out lean/mean journalism.
A smart tactic, but one that still would rely on editorial judgment about what news should be presented. Smells like latency to me.
Yet most organizations that focus on news, and its associated verticals (financial news, entertainment news, sports news, and so) pursue a model that relies upon editorial judgment to determine what gets published. This has remained more than unchanged, calcified, perhaps, since 1994. It is the most linear adaptation of an old-school publishing metaphor that one could concoct.
The other technique that news sites can pursue is social media sensing about what people deem important. This doesn’t suggest that humans can be done away with in the story selection process, but determining weight and heft, and what is important, should be driven by what people are paying attention to right now, not editoral-historical perspective about importance.
This can be done by sampling Twitter, Collecta, OneRiot and other real-time engines for the topics, stories and subject matters at hand, and determining the relative weight of story importance based on mentions, velocity of the mentions, the authority of the person and/or the site making the mention, and other algorithmically-driven analysis.
The Huffington Post has managed its way to 24MM monthly uniques using a small group of freelancers, some staff, and better judgment about the zeitgeist by leveraging social trend data to determine subject matter. This is higher order prioritization than human judgment, but still trails real-time by a lengthy period. They also have a right-column that shows their most popular stories, and this serves to drive a great deal of traffic on the site.
Even faster in terms of approaching real time is the approach taken by Jonah Perreti’s smart BuzzFeed. Compete places them at 2.2MM monthly uniques with a network reach just south of 80MM monthly uniques. They know, more than most other sites, when a piece of content is going viral, which is different than news judgement, but my point is that it is complementary, and worth as much (and requires fewer people to inject themselves into the process.
Most news outlets are using humans for editorial judgment. The next level in the just-memed “Pyramid of Publishing†is the HuffPo, a little bit of the Daily Beast, and some elements of BuzzFeed that are news-related. There are others in that next tier, to be clear.
The top of the pyramid is rarified land, with a few enterprising souls and some big companies dabbling in real-time + news categories +/- media assets.
Some examples include Brizzly, Bitly TV, the aforementioned Collecta and SkyGrid (iPad only.) (FanFeedr, does this on a personalized basis, for sports.)
There is going to be ongoing friction between these five categories of news prioritization:
- Editorial: fleshy people making decisions about what is important
- Real-time: streams of relevant news around a topic
- Wisdom of the crowds: leveraging implicit and explicit interactions, sharing, commenting, and views, to determine popularity.
- SEO-driven: Demand Media, Aol and others are pursuing this approach
- Local: This is actually an orthogonal dimension that can frame any of the above items on a geo-located basis.
For organizations that want to carve cost out of the equation and still drive customer value, the prioritization should be Real-Time, with a “What’s Hot†component. For extra credit, provide a local window into the above items.
The new Newsweek, despite its shortcomings (and lack of time-stamp on its articles), is broadly tackling this head-on, which is a good thing, and others who pursue this strategy will find greater dividends on their news-gathering investment.
For another point of view on this subject, Robert Scoble has posted a smart take on the seven needs of real-time curators here.
Geoff, you are quite right, but understand that I am not advocating an end to journalists, but suggesting that the EDITORIAL process is in need of reconsideration. With few exceptions, real people, with not small talent, need to actually write material. My thesis is that we need fewer/less/no humans to decide what is important or what other people should see. As the folks say at Google, we are in violent agreement on your point.
Thank you for reading the piece, Ty
Great post and clearly you've thought through this more than I. Agree that real-time curation can really drive pageviews – and it is pageviews that these publications still depend on for their revenue right? Biggest question I have is how do you start the viral ball rolling? These pubs – Brizzly, Bitly TV, Collecta…FanFeedr can surface and virally spread stories across the social graph – but where does the original reporting come from? You're back to the fleshy journalists and freelancer's that are necessary to do the primary research and have a solid POV on a story right? Or is the real biz opp to simplly do a better idea at curation and build an audience without the expense of ANY staff?