Last fall, I wrote how the trend of Big Data – huge, almost unmanageable volumes of data that large enterprises collect via social media, internet, mobile devices, sensing technologies and other means, is impacting data storage trends in the SMB sector (i.e., a likely increase in cloud-based storage among SMBs). However, once that Big Data is stored, SMBs need to analyze it in order to obtain ROI (return on investment) on increased data storage investments. And MSPs just may be able to help them out.

According to a new survey of more than 800 SMBs by research group Techaisle, about one in three (34 percent) of SMBs is interested in Big Data. Of those companies, about three in four (73 percent) prefer to use the open source Apache Hadoop software development framework, which is designed to support distributed processing of large data sets.

In addition, the survey shows 40 percent of SMBs are interested in predictive analytics, largely driven by marketing departments that are looking for ways to cut costs, improve profitability, and thinking of ways to harness the data that they are generating within their businesses. However, Techaisle data suggests the lack of expertise within internal IT is becoming a main barrier for Big Data analytics adoption, along with a lack of use cases and clearcut path to ROI.

MSPs Have Several Big Data Options

MSPs serving SMB clients have several options to pursue relating to Big Data analytics. Techaisle analysis concludes that much like with Big Data storage, SMBs will most likely turn to cloud-based solutions as a service for Big Data analytics. Certainly, the scalability and relative affordability of the cloud will make it an attractive option for SMBs trying to derive insights from their caches of Big Data, and SMBs should be sure to have this option available. But there are other steps MSPs should take, as well.

Most importantly, MSPs seeking to develop an SMB Big Data practice should become intimately familiar with Apache Hadoop if they aren't already. Hadoop has become the de facto framework for supporting all manner of Big Data activities. In addition, MSPs should become familiar with other software and infrastructure options that exist for managing and analyzing Big Data, such as RDBMS/SQL and NoSQL data management systems, as Hadoop is not optimal for every Big Data situation (for example, sets of data that change frequently, such as transaction data). MSPs and their SMB clients should remember that Big Data is a big opportunity, and does not have single narrow solution.