Behind the Netflix Debacle, A Core Data-Management Blunder

By Jaime Fitzgerald

What was Reed Hastings thinking? After decades building his reputation as one of the most data-savvy, customer centric, and innovative corporate leaders of our generation, the Netflix CEO stumbled badly when he announced plans to split Netflix operations into two separate businesses. After a barrage of criticism from customers, stockholders and analysts, Netflix announced Monday that it will abandon the ill-fated plan.

What About Customer Experience and Data Management? Netflix seems to have failed to consider the customer experience they were creating…and the role that data management, technology, and process best practices play in creating that experience.  They must have been so focused on the strategic aspiration of focusing on fast-growing streaming video that they forgot the basics…

To put it bluntly, their plans would have intentionally created data silos (something most savvy companies strive to eliminate!) and as a result would have reduced benefits to customers from existing data assets, technology, and processes.  Ironically, the same company that sponsored the pioneering Netflix Challenge—effectively crowd-sourcing the optimization of their movie recommendation model–was planning to intentionally create a less integrated, less customer-friendly version of itself.

Replacing a single service with two. Netflix had announced that it would “spin off” the DVD portion of its operations into a separate business called “Qwikster,” which would be accessed via a separate website, billed separately, and operated independently.  Netflix subscribers would access streaming video only.  Related price changes would increase the combined cost of subscribing to both the DVD and video services.

Data Silos:  a Sure-Fire Path to Bad Customer Experiences.  By splitting a single service into two, the Netflix plan would have made existing data assets less valuable by creating two silos.  On one website, customers would have access to the library of  available DVDs.  On a second website, they could select on-demand video content.  While most firms are striving to integrate their data, the Netflix plan would have “ripped apart” an existing data ecosystem of significant value

Less Customer Value via Data Silos.  By creating two operational silos, with associated data silos, Netflix would have degraded both of the two main drivers of customer value:

  1. Access to videos The primary reasons customers subscribe to Netflix is to watch movies and TV shows…how content is delivered is a secondary consideration.  Replacing a a single subscription to the complete library of available titles with two subscriptions to partial libraries would degrade the customer value proposition.  Looking at this from an information management perspective, this is the equivalent of splitting a single “product master” into two databases…
  2. Customer convenience.  Aside from watching videos, Netflix delivers a major customer-experience benefit through convenience.  Customers are able to easily search and find movies they want to watch (actively and via predictive recommendations!) and add these movies to their Queue.  By splitting their service into two parts, Netflix would have reduced this benefit also.  According to the Wall Street Journal, customers would have needed to log into two websites to search the library of available titles…something they can presently do via a single portal:

“The split-up plan would have meant Qwikster would handle billing for the DVDs-by-mail service while Netflix would bill for the online viewing. Their two websites wouldn’t have been linked, meaning customers would have to check both sites to see whether a movie is available on DVD or for online viewing. Viewing recommendations for one wouldn’t factor in movies from the other operation.” (read the whole article here)

Ironically, Reed Hastings and the Netflix team may have lost sight of basic principles of data management, customer experience, and process optimization because they thought they had “mastered” these domains.  Their fall from grace reminds us how important it is to protect, maintain, and enhance these core enablers of business results.

Related Posts Plugin for WordPress, Blogger...

  12 comments for “Behind the Netflix Debacle, A Core Data-Management Blunder

  1. October 17, 2011 at 8:48 am

    In the Age of the Platform, it’s all about building a seamless, integrated experience. Multiple queues would have portended Netflix’s death, especially with Amazon, Apple, and Google chomping at the bit.

    • Jaime Fitzgerald
      November 14, 2011 at 2:24 pm

      Thanks Phil for your comment! I agree, with the competition using information and technology for all it is worth, Netflix needs to extract all the value they can from their franchise, their brand, and their data.

  2. October 17, 2011 at 4:29 pm

    Your comments on the user experience are well taken. Certainly navigating between two websites would have been cumbersome. Obviously a ton of people agreed. But that’s a user experience design flaw, right?

    I think the fundamental data management issue revolves more around the back-end organization of customer data. Are you also saying they were going to split the data architecture into two physically separate systems? That may well have led to problems. But I haven’t seen anywhere that this was being proposed. Did I miss something?

    And there is no reason why a single data architecture could not have supported both user experiences. Many larger organizations support multiple lines of businesses through a single data architecture.

    I believe data management must accommodate complex business missions, however wrong-minded they may appear in hindsight. But these sorts of business scenarios are not necessarily wrong-minded merely because they are complex. Isn’t it our job as data professionals to determine elegant ways to deal with these complexities.

    • Jaime Fitzgerald
      November 14, 2011 at 2:16 pm

      Hi Max,

      Thanks for your comment on the post. I agree with your points, and am glad you shared them. I’ll try to put them into the context of what I learned why researching and analyzing the situation for the post….

      As you said, a single data architecture CAN support a variety of user interfaces. And Netflix never commented on how they planned to adjust back-end architecture related to the split.

      I share your perspective that for a given business objective (in this case, Netflix’s goal of splitting their business), the job of data professionals is to design the most elegant solution to support that goal. My objection is not about the way Netflix manages data per-se (I don’t have visibility into that) but to the way Netflix executives lost sight of the basic principles of what makes data valuable (improving customer experience, delivering customer value, usually through optimal use of integrated data)

      To your question about what Netflix indicated about their plans for data: Netflix described their plan to split the business-and the front-end interfaces–by limiting the data each business would “run on.” Gone would be the “single product master” consisting of both DVD and streaming titles. In it’s place: two product masters, one for digital, one for physical DVDs. The same split would occur–according to reporting by the Wall Street Journal and Others–for the recommendation engine.

      Thank you again for your comment. I look forward to further dialogue. Cheers,

      Jaime

  3. Glenn Thomas
    October 21, 2011 at 8:34 am

    Jaime,

    I believe the current attitude of industrial greed may have overpowered their powers to think logically about what their customers actually want.

    • Jaime Fitzgerald
      November 14, 2011 at 2:25 pm

      Glenn,

      Thanks for your comment. I agree, it seems to me that they “forgot what got them where they are”…which included great strategy, and in many cases, great strategic use of their unique data assets.

      Thanks for commenting!

      Jaime

  4. October 21, 2011 at 3:02 pm

    Undoubtedly a bad plan from the beginning, and you make some great points here, Jaime. Nice job.

    I think what was driving this is that one part of their business (DVD delivery) will eventually become obsolete, as the physical shipment of bits around on a DVD via the US Postal Service makes less and less sense in a converged, fiber-to-the-home environment. And the other part (streaming video) is a high growth, high potential business.

    They were going for the clean separation, so that one could grow quickly without the other one pulling it down.

    But no question, it would have been a bad move for customer experience, and hence for the business. I think they’ve done the right thing by walking this one back.

    • Jaime Fitzgerald
      November 14, 2011 at 2:33 pm

      Dan,

      Thanks much for your comment! I agree, there WAS a strategic rationale for the changes…and yet, they didn’t think it through with sufficient care. One of their key advantages was convenience, and even customer intimacy via their recommendation engine….both built on integrated data. When they made this change, they didn’t think about some basic better practices with the customer in mind.

      That said, to your point they were (relatively) quick to roll back they mistake. That impresses me and I am rooting for them to survive and excel.

      Thanks for your thoughtful comment!

      Jaime

  5. Ron Wolf
    November 7, 2011 at 1:53 pm

    Mr. Hastings hS and continues to cash in $1M per week in Netflix stock all thru this debacle.

    A data problem? Come on, that’s a stretch. This is yet another corporate accountability problem.

    • Jaime Fitzgerald
      November 14, 2011 at 2:01 pm

      Thanks for your comment, Ron. Point well taken, there may be more than one problem here… -JF

  6. December 13, 2011 at 4:29 am

    I do agree with all the ideas you’ve offered on your post. They’re very convincing and will definitely work. Nonetheless, the posts are too short for novices. Could you please extend them a bit from next time? Thank you for the post.

    • Jaime Fitzgerald
      December 23, 2011 at 11:45 am

      Hi qnet,

      Thanks for your comment. So to confirm that I understood, you would prefer a longer follow up with more details regarding the ideas and conclusions?

      Happy holidays,

      Jaime

Leave a Reply

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