House of Cards

House of Cards logo
I recently heard a great news story on National Public Radio on my way home from work that intrigued me and I thought I’d share with my one reader.  Specifically this article talked about metrics used by the hit Netflix show House of Cards.  Netflix is a great company to work for if you are a data nerd as they have extremely granular data not only of their customer demographics, but also, and more importantly, of their behavior and treatment of your product.

You’d think an organization with that much data, and a well-earned reputation for data savvy management, would pass those insights on to their production and development teams so they would know the best loved characters, how plot lines were received, and other critical information.  So how much does Beau Willimon, the show’s producer and creative driver, see?  Well, in his words: “I know virtually nothing.”

Wait, how does that happen?  This is Netflix, one of the major companies in the middle of the big data revolution and their hottest product right now is being run by a guy who doesn’t see any of that data?  This flies in the face of everything we know and love about data!  Aren’t we supposed to be running our business off of every data stream we can find?

I loved this story because it illuminates some very important facts about big data and how it should be actually used.  The first is that the data itself can be both a great tool…and a great detriment if you’re not careful.  As the interview continues, Mr. Willimon explains that he doesn’t want any of that data either.  He doesn’t want the distraction for his team of creative types and that kind of detail would have them second guessing each and every creative decision that they made.

Interestingly enough, we see that Netflix made this same decision.  They guaranteed Beau an extremely generous run of 26 episodes without seeing the first one.  By giving Mr. Willimon that promise, he was able to develop a rich canvas of characters and plots that don’t feel rushed or extreme.  The bread was given time to rise!  In contrast, new shows often have extremely short leashes with many being cut long before they complete their first season based on a few poor ratings.  The ability to focus on the product and not the numbers gave his team the creative freedom and also the peace to do the best job they could, without looking over their shoulders constantly.  They didn’t have to pander to their customers (us) with every episode but instead could, as master craftsmen, make the best product they could.

What Netflix understands is that the data itself is a representation at best of customer behavior, not of the quality of the product itself.  So as they created this series, they needed to focus on making a great product, independent of anything else.  Measuring the customer’s behavior and treatment of the product is important to know when choosing new products, marketing, and other activities; but you simply can’t try to create the product itself based on these data points.  Could you imagine Leonardo da Vinci stopping every paint stroke, to get the approval of and seek direction for his next stroke, by a gallery of people behind him?  What a mess the Mona Lisa would be!

We must never forget that people aren’t machines, constant, unending feedback loops dampen creativity and destroys vision and strategy.  While we should use these tools and technologies to shape behavior and help people maximize their limited resources, we shouldn’t get lost in the fact that they are indeed people who need room to be creative and experiment.  We often complain about reports and data being ignored by management when making decisions, that’s not always a bad thing.

The trick to a successful business is to find the right balance between usage of the data and quality decision making by a focused and knowledgeable management team.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>