A few months ago Karl Flinders at Computer Weekly asked whether systems integrators were slowing the adoption of open source in the UK government. Flinders wrote about how even though open source would seem like an attractive option for cash strapped agencies, the adoption has been surprisingly low. That could be due to systems integrators whose revenue depends on selling proprietary licenses to customers.

Some have assumed that systems integrators would be early adopters of open source, as it could give competitive advantage to firms by allowing them to cut licensing costs. But according to a 2009 survey by Jaspersoft, the truth is that consultants have lagged behind IT departments in open source adoption.

But look at big data. It’s not just changing business as usual by providing new insights. The need to run big data applications is changing business as usual by necessitating the adoption of new approaches like data science and DevOps, and it’s necessitating the adoption of open source tools that don’t have proprietary alternatives.

There is no real direct proprietary alternative to Hadoop. Microsoft started building one, reached beta and then killed it off in favor of supporting Hadoop. LexisNexis decided to open source its own Hadoop project. There are proprietary Hadoop tools from companies like Cloudera and MapR, but there isn’t a direct proprietary alternative. Columnar databases from companies like HP Vertica and Teradata can handle big data up to a point, but they’re not a direct replacement for Hadoop.

There are plenty of other examples. Apache Cassandra came out of Facebook, Membase came out of Zynga and Flock came out of Twitter. These aren’t open source clones of proprietary commercial software (Cassandra is inspired by Amazon Dynamo, which isn’t available commercially). Each is a purpose built tool that filled a role for the company that built it and is now open source. Their primary competitors are other open source tools like Riak, Redis and Neo4j.

Service providers will be an important part of big data in the enterprise. For example, Look at Cloudera’s strategic partnerships with companies like SGI. While much of the traditional service providers work might be affected by shifts towards SaaS and simple data integration tools, there will be new opportunities in helping enterprises leverage open source big data, data science and operations automation tools.