Abstracting and recombining all the way to the bank

In the past I’ve never understood what people really mean they say “think outside the box” but Jim Harris, in a recent OCDQ blog post, helped me figure it out.

Mr. Harris ends with this provocative line: “the bottom line is Google and Facebook have socialized data in order to capitalize data as a true corporate asset.”  The post starts with a cold war analogy and proceeds to describe how Facebook and Google have made big money as “internet advertising agencies:” offering free services with which users (like us) serve up personal data in return for use of the service, then selling advertising space based on our data (hopefully anonymized).

Mr. Harris’s analysis shows the power of what data modelers and object oriented designers call abstraction: Google and Facebook ostensibly provide products like Gmail to customers like you and me but they really don’t do that at all.  They make services like Gmail available to users, but their real product is the data those users provide, and their real customers are those who pay for ads presented by Gmail, facebook.com, and similar products. In some cases, like Google Places, one suspects that the user is both paying advertiser and data provider.

In data modeling, abstraction might be defined as separating out commonalities between similar but distinct concepts into a common entity.  Both customers and advertisers in our example have telephone numbers, email addresses, physical addresses, and names.  A data modeler might put those attributes into a separate entity, typically called “Party”, and design the database so that the Party table would contain a row for each customer and each advertiser, where each row would represent one or the other.  Such a setup might improve business process consistency, since both customer and advertiser contact information could be handled the same way.  Of course, it might be disadvantageous if the business rules require the two to be treated differently.

However, Mr. Harris’s post offers a much more significant advantage of abstraction.  Abstracting business concepts makes it possible to think about them in new ways. Picture an early 2000’s Google or Facebook strategic planning sessions: “what would we do with it if we had data about 100 million people, with their contact information, email or “wall” communications, and search patterns”. Even now I bet they are just beginning to mine the possibilities of that “out of the box” conversation.

Abstraction can be a powerful tool even for the rest of us who aren’t trailblazing internet innovators.  It is easy to come up with negative examples, where being too specific causes business inefficiencies:

  • One Fortune 500 company left contractors out of its personnel systems, making it difficult to standardize on-boarding processes including security training and administration, which operated differently for each department.
  • An agricultural company had a finished goods system that custom-built products for each order, so each order needed new programming to handle the “new” product.  That company could have saved big money in app dev avoidance alone if their system varied the product’s attributes across different dimensions rather than seeing each order as a brand new product.

Not to dwell on the negative though: here are some success stories, previously cited in an article I wrote with Tri Truong, on how organizations have abstracted business concepts and recombined data about them to improve business performance:

  • During the mid-’00’s Lowe’s gained a competitive edge with “data-driven shelf plans,” which objectively identified which [products] generate the most profit, where they get the best attention and what season gets the most action. The results are run through a series of rules…[and] out comes a floor and shelf plan.” Presumably this program ignored distinctions a human might apply to co-locate unlikely products for better sales.
  • In another highly publicized example, Capital One emerged in the mid-90s as a Fortune 500 company through an analytical, data-centric approach that helped grab credit card market share and slash its costs. One of the key insights that drove that growth was recognition of an “underserved” market for revolving credit among those with low credit scores.  Rather than dividing the population between qualified and unqualified credit applicants, Capital One revolutionized the credit market by thinking of everyone as qualified and modifying the product to fit the customer.

So when people talk about “thinking outside the box”, I think it means taking business concepts up to the next level and recombining them in new and promising ways for fun and profit!

Leave a Reply

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