I was reading an interesting article in Forbes titled “Hortonworks Wants To Own Big Data Without Owning Anything“.  This explains some of the history and business strategy of Hortonworks in the Hadoop/Big Data space.  The article also contrasts the business strategies of Hortonworks and Cloudera.  I should also point out that the other pure play Hadoop company – MapR – should also be included as it too has an interesting strategy.

Simba has been involved in the Big Data and Hadoop space for almost three years now.  Our business is to connect applications and users with data.  As such, pretty much any data source needs our technology to build the drivers, providers, and connectors that open up their data to common business applications like Excel, Crystal Reports, Tableau, etc.  Because of the unique position we are in, we know early on how the different data sources are doing and also which are the most popular applications that people use.  In the Big Data space, we have seen the evolution of Hadoop and NoSQL.  We saw early on that SQL was going to be important for both Hadoop and NoSQL.  The reason for this is that SQL is the lingua franca for connecting commonly used applications to any types of data.  Even when dealing with what is commonly called unstructured data, people forget that for data to have value, there must be some form of structure.  The structure may not be explicit like traditional SQL sources, but often there is some implied structure like in JSON documents.

When we look at Hortonworks, Cloudera, and MapR, I see all of them have great ideas of how to take the data stored in Hadoop and expose it for people to use and gain real value.  They are all investing heavily in things like Stinger/Tez, Impala, and Drill.  These are the investments that look very interesting going forward and it will be an area to watch as we head into 2014.