The role of innovation in news has come up in several conversations I’ve had with folks over the last few weeks, and I’ve come to the conclusion that the pursuit of innovation may be fun as all get out, but on its own it does not do enough to move the industry forward. What we need instead of innovation is experimentation.
What’s the difference between innovation and experimentation? Innovation only values success. Experimentation also values failure.
Innovation is about creativity alone. Nothing wrong with that. You can’t have experimentation without it. I’ve promoted a lot of ideas at news organizations simply because they were different or because nobody else was doing them. Innovation is fun. Probably hundreds of online journalists — those that have even an ounce of self-awareness anyway — have uttered the words “I really like building stuff. Then ‘they’ can maintain it once we have it up and running.” Too often innovation in newsrooms is not about throwing spaghetti at the wall and seeing what sticks, and never wondering why it sticks. Is it the wall that’s sticky or the sauce or the noodles?
If we stop at innovation, the only thing we will ever learn is that under certain conditions something can be done. You can stream video over the Internet. You can create an interactive NCAA basketball bracket contest. You can use Ruby on Rails to develop a microblogging social network. You can fly nonstop across the Atlantic. You can climb Mount Everest. You can put a man on the moon. Cool… but can *I* do it? And if so, why should I?
Those are the questions that experimentation answers.
“For an experiment to be useful, you should know in advance the criteria for success or failure,” said Phil Meyer, the author of Precision Journalism. “Even better, you should have action standards in mind, e.g. if it works, what would you do differently because of that knowledge?
I want to highlight two things that he said:
1. You should know *in advance* the standards for success or failure.
2. The outcome of the experiment should cause you to behave differently.
Let me tell you two personal stories about how innovations in which I’ve been involved have stopped short of experimentation, and why they would have been better if they had not. The first story is about the now common “most popular stories” feature on news Web sites. The second is about multimedia.
Several years ago, innovative news Web sites began to post lists of the “most read” articles on their homepages. Cool. First, that’s just technically interesting and that instant and automated audience feedback loop to itself is something you just can’t do in print or broadcast. Second, it just feels democratic and — to me anyway — socially groovy.
When I started having conversations with folks in my newsroom at the time about whether we should do this, it began as a religious battle — one sect arguing for the autonomy of the editor over the tyranny of the masses, and the other arguing for reader empowerment and innovation. The smartest folks in the newsroom often argued both sides with equal skill.
First experimental opportunity missed: Is posting “most read stories” to your homepage more democratic in some way? Do readers perceive it as a value, either by their behavior or their opinion? Under what circumstances will the editor-driven headlines and the reader driven headlines be different?
A pretty quick scan of site logs led most folks in the newsroom to agree with the hypothesis that there were some pretty interesting differences between the editor-driven and reader-driven headlines. But I still don’t think we really know very much at all about what those differences might be.
The next debate was about how to define “most popular”? How often should we sample? And when we did sample, what should we measure? Cumulative popularity or the change between two different sampling periods? Page views or unique visitors?
Second missed opportunity: Looking for differences in what those different measuring techniques would yield. And if they did yield a difference, how would we behave differently?
The most interesting debate we had before launching was the most interesting to me — the idea that it rewarding the “most emailed” articles would be closer to our editorial mission that rewarding the “most viewed.” In the age of online journalism, I suspect that the greatest fear of newspaper journalists is that they don’t become their own perception of television journalism — a ratings driven popularity contest that favors crime and cats because that’s what is “most viewed.”
There was a notion among many people — and I was in this camp — that it was “better” to reward our more engaged and influential readers by posting the articles that had been emailed the most. This was sort of the digital version of promoting the “watercooler stories” — the pieces that readers would discuss around the watercooler at work.
Of course, we knew that we could only count the people who emailed stories using our site’s button to “email this article” and we knew that most of us didn’t use that feature ourselves, favoring instead the simple act of cutting and pasting a link in to a separate email application that already had the addresses of all our contacts. But we didn’t know how many “votes” we might be discarding and we didn’t know whether that difference was significant in any way.
But the real crime here is not that we didn’t launch this feature with a mind toward experimentation and discovering a broadly applicable theory that would change our behavior. The real crime is that we still don’t have the answers to these questions. Or at least I don’t. I don’t think we have any idea whether or how this now ubiquitous feature has changed audience behavior, newsroom behavior or advertiser behavior.
Around the same time this debate was happening in newsrooms, another debate was developing. The use of audio, video and user-driven animation to tell stories on news sites had led to an explosion of journalistic creativity. It had brought to newspaper newsrooms people who truly had a different way of looking at journalism. Their presence alone was inherently innovative, and the stories they developed were widely reward by their peers around the globe.
But their stories were often incredibly expensive and often received very limited audiences. The multimedia staff rightly argued that the lack of audience was because they didn’t receive “enough” promotion. At the same time others were calling for managers — both editorial and business — to clearly and publicly describe to the staff how they defined “success” of these projects in which we were investing so heavily at the perceived exclusion of other priorities in the newsroom. In a world of scarce and diminishing resources, their rationale — “innovation” — was not enough. Mostly, it was not fair to the multimedia staff that was being pulled apart by a demand for innovation and a demand for accountability from a variety of angles.
What we needed — and still need — is experimentation. We have no qualitative categorization of multimedia content and we have no quantitative measures of the effects that these different types of content have on the audience, the advertisers and the newsroom. Most newsrooms have no clear defintion of success or failure of the most innovative multimedia projects in their newsrooms. Does that mean they should stop doing them? No.
The Value of Failure
There has been a lot of discussion recently about the value of learning how to “fail fast, and fail cheap.” Unfortunately, most news organizations fail slow and expensive. Not your newsroom, right? Well, ask yourself this — how many staff hours were spent on the last redesign of our site? And does at least 90 percent of the newsroom know and agree on the lessons we learned from the changes we made? And has management provided leadership about changes in direction based on those lessons. If you answer no to any of those questions, then your newsroom is failing in the wrong way.
Banning the word “innovation” from your newsroom and replacing it with a culture of “experimentation,” doesn’t guarantee that you will fail fast and cheap. But it does make it much more likely that you will at least know when you’ve failed and when you’ve succeeded and it will reduce the opportunity cost of failure because — at the very least — you will learn something. You’re more likely to be like Mae West, who when faced with the choice between two evils always picked the one she never tried.
And Mae West was sexxy! You want to be sexxy, don’t you?