Editor’s Note: This is the second of two guest posts by Scott Porad, who is part of the team behind I Can Has Cheezburger? and the Cheezburger Network, the leading online destination for user-generated humor and entertainment content.
Previously I addressed the misconception that user-generated content is free. To make user-generated content work, Cheezburger expends significant cost to sift through all the user submissions to find the best quality content. However, including this expense, content costs us less to acquire and is undoubtedly of higher quality. This fundamental win-win is the promise of crowd-sourcing and user-generated content.
With that in mind, what can news journalism learn from I Can Has Cheezburger? There are probably many lessons, but one that stands out to me is a fundamental shift in the concept of reporting from “sourcing” toward “filtering.”
As an outsider, someone who never has worked as a professional news reporter, it seems that in the pre-Internet era the primary constraint on a journalist was a lack of sources with quality information. That is, in order to know what was happening somewhere, you had to know someone there who could be a source. After finding a source, a reporter would verify the quality of their information, typically by corroborating it with additional sources. In short, the name of the game was finding and developing sources.
Today, the problem is the opposite: a journalist has a million sources–anyone with e-mail, a cellphone camera, a blog or a Twitter account is a potential source. The issue now is not finding sources, but figuring out who among the myriad of them is actually providing the right information.
By way of example, let’s refer to the recent “man overboard” report on a Washington State Ferry. Within moments, Twitter was alive with real-time reports. All the news reporters had a million sources, including the US Coast Guard which was tweeting events as they happened.
But that didn’t mean they were all quality sources: yes, a report was made, helicopters were dispatched, and a search was on. But, there was no man overboard; it was a false alarm. Suddenly, the job of the news reporter changed; no longer was it “where or how can I find someone who will tell me what’s going on aboard the ferry?” Now it was: “of all these people reporting what’s happening on the ferry, who is telling the truth?”
Earlier this year I was at a panel discussion that included two leading online technology reporters: Louis Grey of louisgrey.com and Marshall Kirkpartick of ReadWriteWeb.com. Each reporter discussed the tools and techniques they used to sift through the endless volumes of sources online in order to reveal the right stories to be reporting.
I was struck as Kirkpatrick described how, at considerable and non-trivial expense, ReadWriteWeb analyzes their sources to understand which of them are the best. Like Cheezburger does with moderating images and video, ReadWriteWeb performed this analysis using a human-based process–it was not automated with computers. And, like Cheezburger, the name of the game for ReadWriteWeb is being able to effectively sift through the noise to find the signal, in a word: filtering.
And, a final example to illustrate the importance of filtering, from beyond the media world and into e-commerce: Amazon.com has relied on user-generated content since day one with their Customer Reviews. However, popular titles often have thousands of reviews submitted. For example the most recent Harry Potter book currently has 3,328 user-submitted reviews. Each of these reviews is reviewed by an Amazon employee for spam, language, appropriateness, etc.
The problem for Amazon, like Cheezburger and ReadWriteWeb, is not an issue of having enough content, but rather an issue of filtering out the best. Like Cheezburger, they have solved this problem through a combination of employee and user moderation. Leading usability expert Jared Spool estimates the solution has been worth $2.7 billion to the company.
Additionally, the reason companies invest in processes that filter for quality is because better content drives growth. Higher quality content equals positive word-of-mouth equals more traffic and users. Of course, more users leads to more submissions leads to better content. The lesson is that content publishers can’t simply publish every piece of user-generated content that is submitted, or it will diminish the quality of the product resulting in the exact opposite effect.
There is still value for news reporters and organizations to be “source leaders.” I am one who believes that there is an important role for the professional news reporter in our society. But it is clear to me that user-generated content or crowd-sourced information is a valuable addition to news journalism because it can yield more and better information, and often faster.
However, as I’ve illustrated, that information comes with a cost which is finding ways to separate the signal from the noise. The task news reporting has shifted toward filtering. My view is that winners in the Internet era of news journalism will be the people and companies who, like Cheezburger, ReadWriteWeb and Amazon, develop systematic ways of filtering the flood of user-generated content and sources down to those with the best content. The result will be higher quality news and information, that is more relevant and on target with the audience, at a lower cost.
Scott Porad is the part of the team behind I Can Has Cheezburger? and the Cheezburger Network, the leading online destination for user-generated humor and entertainment content. Scott writes about web startups, new media and miscelleany at http://scottporad.com, and you can follow him at http://twitter.com/scottporad.
Mark Briggs 
Posted in 
Journalism has always been about selection and evaluation of sources; this is nothing new.
What a journalist needed to do is to make sense of the current story and related, previous stories, to have it all make sense to the reader.
For example, in the WSF example from above, repeating selected tweets is like copying the daily police blotter. At some point a journalist has to ask and answer questions:
* How is the report of a person overboard made on a WSF?
* How often are there false reports of a person overboard?
* How does the report of a person overboard affect those riding the ferry?
* How does a WSF respond to the report of a person overboard? Is the Coast Guard typically involved?
I’m sure the reason why cub reporters used to start out by working the police blotter was to educate the reporter to what was happening in the city. Only then could questions be asked for journalism.
A stenographer can repeat a story. It takes a journalist to place it into context and make it understandable.
Porad is right – our job is shifting from sourcing to sifting.
News media seem yet to pick up on this trend. Makes me think we need to re-examine old concepts of cyberjournalism and update them for today’s market.
Instead of relying mainly on sourced news, old media needs to place equal emphasise on sifted news – i.e. follow where the crowds are.
Old media pretty much ignores events like the man overboard, dismissing it as another false lead, rather than an opportunity for critiquing cyber news and providing context.
Good point on the cost aspects of seperating the signal from the noise. UGC isn’t really “free”.
On another note, we need the traditional reporter simply because they do the tough work that UGC contributors generally don’t.
Generally, it is the old-fashioned reporter who investigates corruption in Washington, who covers the war in Iraq, who risks his/her life in Afghanisthan, who investigates child abuse in Rwanda etc.
Of course, there are exceptions – I’ve written about the “twitter revolution” in Moldova, Iran –
However, for the most part, it does seem like we need reporters to do the really tough work.
Porad is only right up to a point. Firstly, the user-generated model only works for those parts of the world and those stories where the players or bystanders are networked, or where the story is being played out in a networked space. This will not work in remote places, for example. It will work less well where the players don’t speak English, because this places a whole new set of burdens on your ‘filter’. There are many other examples one might think of.
Secondly, this style of filtering of UGC is a passive form of journalism. It waits for events (and only that fraction of events that are ‘networked’ as above) and then starts filtering. It doesn’t seek out news. It waits for what the user thinks is newsworthy. That’s a big hole in the traditional model of journalism.
Thirdly, this style of ‘journalism’ won’t work for stories that aren’t being played out in real time in some networked space. It won’t work for discussion of climate change policy, for example. That requires traditional research and interviewng skills before the the journalist can then apply their filtering skills to the technical information of the experts. This is not something that you can twitter.
Every example of how Twitter, etc., is supposedly changing reporting seems to rely on EXTREMELY RARE tragedies, disasters or sensations.
I don’t know about your hometown paper, but in the one I work for, almost all of what you’d call “breaking news” (aside from the sports and arts coverage) consists of
a) Cops and courts — situations known only to a tiny group of private, deeply interested and unreliable individuals.
b) Political actions — city and county governments doing stuff, known only to a handful of deeply interested and unreliable people present.
c) Studies, findings and reports — released by governments, nonprofits and businesses.
d) Pseudo-events — announced press conferences, etc.
It’s hard to imagine online-only reporting being at all reliable in (a) or (b), and it’s hard to imagine it being much more effective in (c) or (d) than simply picking up the phone, firing up the Internet or going to the damn press event.
In situations like document dumps or earnings reports, many eyes can be an effective way of finding hidden gems or coming up with provocative questions.
But in almost every other local reporting situation, what’s really happening is that a reporter is synthesizing facts that haven’t been brought to light or widely shared, and explaining why they’re important. This is not a task that crowds do well. Unless they’re aided by large software programs, it’s not a task crowds do at all.
Anybody who thinks floods, fires and ferry accidents are what local reporting is all about needs to look more often at his or her local newspaper.
I think the comments above demonstrates that there will always remain a case for a hyperlocal patch reporter.
While curating will become an increasingly important skill for a journalist, as Jerry says, there is a level at which there simply isn’t enough connected individuals online to generate the information stream through which to sift.
In the city in which I live in England, there is a growing network of people out there, but it’s simply not strong enough to rely upon in creating stories.
And as Michael says, the journalist will still have to rely upon digging out and presenting stories regardless of what opportunities are presented using new technology.