email iconemail phone iconcall
Scott Hirleman

Big Data or Big Hype or Why NoSQL Matters

By Scott HirlemanMay 9, 2014

With all the marketing hype surrounding Big Data and how it is ‘the most important IT trend this millennium’, what is lost in the noise is the significant impact Big Data technologies can have on real companies especially around customer interaction. Big data truly can be transformational, but its focus needs to be on what problems it can actually solve AND what can be done with these big data technologies that couldn’t be done before. Solving current problems is what reactive companies are looking to do while proactive companies are looking at new technologies to enable them to interact with their customers in new and exciting ways, generating the actionable data needed to beat the competition.

NoSQL is one of the cornerstone groups of technologies of big data and is leading the new breed of databases addressing challenges that plague traditional legacy systems. NoSQL databases were created to solve real problems that relational databases management systems (RDBMS’s) have major issues with.

  • Relational databases are difficult to scale (sharding isn’t fun)
  • When you can scale a relational database, it is expensive and complex (licensing costs, expensive servers due to scaling up, long development cycles for new features, etc.)
  •  RDBMS’s don’t handle multiple types of data well (cleaning data in transit… messy) and struggle with certain types of data entirely (e.g. unstructured data, blobs of text, sensor data)

As a result, the practicality of taking big data from managing information to creating engaging customer experiences while using legacy relational systems has been a challenge for most organizations, leading companies to seek new alternatives. Viola NoSQL. A great example of the successful adoption of NoSQL is Netflix. Netflix went from a simple DVD rental company run almost entirely on Oracle to one that streams content to millions of customers worldwide, generating well over 1 trillion transactions a day. Netflix tracks every customer interaction (down to every 2 seconds you watch) to reach 2 goals: (1) make sure the user experience keeps getting better and (2) create and secure content users want. Massive hits like ‘House of Cards’ and ‘Orange is the New Black’ are testaments to Netflix getting it done right by leveraging NoSQL.

Netflix is just one example. Hundreds of forward-thinking companies use this approach to redesign how they interact with customers, suppliers, and partners to generate actionable and valuable data that drive their businesses to new heights. These new interactions generate both new types of data and a more varied set of data, all at unprecedented scale, which brings us back to the original challenge- RDBMS’s don’t handle any of those scenarios well.

NoSQL databases thrive in today’s high-volume, high-variety online applications. They enable companies to be more agile, especially while deploying new features; more flexible (store varied and/or complex data types); and able to support scale without high costs and complexity. While the hardware used to support a large-scale Oracle deployment could easily run into the tens of millions of dollars, the same application on a NoSQL solution like Apache Cassandra could easily cost less than $1M.  Add in the cost of software, and the difference grows even more. Moreover, NoSQL gives smaller organizations the ability to deploy applications in-house for under $50K (or in the Cloud for even less ), giving them flexibility and agility to get a product in front of customers.

As companies continue to innovate and find new ways to interact with their customers, creating large amounts and varied types of information, the industry will continue to run into a bit of chicken and egg problem with questions like ‘When do we have problems that cause us to need NoSQL?’ versus ‘What can we do to make our products better by leveraging NoSQL?’. Forward-thinking companies are asking the second question.



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