Two interesting articles of note in the current Harvard Business Review, one, an interview with John (“I’m not Meg Whitman”) Donahoe, the new CEO of eBay, and a fascinating piece about creating a “culture of experimentation” coincide perfectly with my recent talk with Bill Franks, Chief Analytics Officer at Teradata.    Bill* has become an innovation evangelist to the customers of this cutting-edge, “big data” provider, and it just so happens that eBay is one of them. 

It’s true that “innovation” is one of those corporate goals that remain so unassailably “good” that no one is going to argue with them. However, with the rise of Internet-driven and social business, we need folks like Franks to explain how to foster a culture of innovation in a world where data is moving so quickly and so ubiquitously that failure is not only an option, it is desirable.

The Internet is moving so fast that companies cannot afford to inoculate against risk but have to, instead, anticipate it and move forward, making adjustments as quickly as you can.  Donahoe talks about how eBay is doing just that….”by celebrating failure.”  In an example he cites from 2010 where a pricing scheme was not working, “….There was a second-order dynamic that was depressing it, so over the summer we really drilled down to understand what was going on….  That’s a lot of what business is today: the old, ‘Ready, fire, aim.’” 

The good news is that data management, unlike in the old days, allows companies to isolate a sufficient sub-set to create a “walled-garden” for the purposes of experimentation without putting the entire database at risk.  Companies need to make a deliberate embrace of risk, to test all sorts of wild scenarios, and to create teams internally whose job is to experiment, and yes, to fail.  According to SmartDataCollective adviser and leading data management consultant James Taylor, “I think experimentation is very important and that a walled garden is helpful. I also think that companies need to think about embedding test-and-learn, or experimentation, in their operational systems. Many companies have done well with this, in A/B testing for website design for instance, but more need to think in terms of always running business-level A/B testing to see what works in pricing, in marketing, and in customer service. Only a culture of managed experimentation allows you to learn.”

How does Franks see this working in the enterprise?

How does the "walled garden" for data experimentation benefit management's understanding?

Franks: “The main way I see the "walled garden" / sandbox / data lab benefiting management is that they will be able to see a broader range of approaches and problems addressed.  By allowing for faster exploration, you can get more insights and get each insight faster. Having a great finding isn't helpful unless the decision-maker is made aware of the finding in a way that they understand what it is and what they can do with it.  A solid analytical resource should be able to get a good start at compiling the results.  Someone in the leadership role within analytics can then work with someone in a leadership role on the business side to finalize recommendations, how to present, etc.”

What are the recurring feedback loops necessary to keep innovation going? 

Franks: “Main loops are ensuring that the innovations that work are well documented, but also those that don't.  By knowing paths that have been eliminated in the past, it can help immensely in not wasting time in the future. In addition, management needs to continue to offer up new ideas and encourage others to do the same. 

Most organizations have a culture where everyone wants to play it safe and failure is viewed as a huge negative.  In the type of environment that we're talking about here, you have to expect some ideas to fail.  But, you've chosen ideas you think have a shot and you've determined they won't work quickly. So, there’s little loss."

With recent business history littered with the stories of companies — like Nokia, the old AOL, or Pfizer — that have not innovated fast enough in their markets,  the costs of refusing to fail far outweigh the benefits of playing it safe.  Thanks to new developments in data management,  it’s not only fortune that favors those managers who give a free hand to their walled gardeners.

* You can connect directly with Bill Franks on SmartData Collective by visiting his profile here.