From the Apache Web servers to the Linux systems and Android mobile devices they use, businesspeople know open-source, whether they know it or not. But does that mean they're willing to entrust their business intelligence strategies to the vagaries of open-source development? After all, these are the undertakings aimed at sharpening an organization's decision-making prowess.
That's the question we pose in today's Point/Counterpoint debate, at right, in which two experts stand off against one another on the reasonableness of the open-source analytics movement. And, yes, I have used the word "movement" deliberately.
The idea of open-source analytics isn't new, especially in the Web world -- Open Web Analytics and Piwik come to mind, for example. But a growing number of companies have launched in recent years to take on the BI and analytics mainstays with commercialized software rooted in open-source. The companies include:
- Infobright and Ingres, which provide open-source analytics databases.
- Jaspersoft and Pentaho, which provide open-source BI platforms.
- Rapid-I and Revolution Analytics, which provide open-source data mining and predictive analytics.
And that's just on the commercial front. Loads of open-source analytics software products are available for the taking. Among the most popular is R, which provides a variety of graphical and statistical techniques, including clustering, linear, and nonlinear modeling, along with time-series analysis. PSPP, used for statistical analysis of sampled data, is another example of an open-source analytics project.
In his Point piece, Stephen Samild, co-founder of Analyst First, makes the case that spending on commercial analytics software takes away from investing in analytics talent. That's the main thrust of Analyst First, which encourages organizations to place more emphasis on the people who perform, manage, request, and envision analytics than on the tools enabling the work.
Ajay Ohri, an analytics watcher for Decisionstats.com, takes a more cautionary stance on the viability of open-source analytics. It's not that the technology is lacking, he says, because oftentimes it's not. But he asks organizations to weigh technology capability against such must-haves as customer service, and he suggests that commercial vendors have the upper hand on the oh-so-important extras.
From where he sits -- at the CFO desk at Volunteers of America Chesapeake -- Shyam Desigan says open-source analytics is a great option. As Desigan says in his blog, VOA Chesapeake Heads Toward Predictive Analytics, the Open-Source Way, the use of R and Pentaho's open-source BI platform is enabling the nonprofit to pursue a predictive analytics strategy. Open-source analytics software, particularly when combined with low-cost, easy-access BI-related cloud options such as KnowledgeTree's data management software as a service, can help bring any smaller organization into the analytics fold, he says.
But what of larger enterprises? Can a larger, more traditional company benefit from the use of open-source analytics? Read the debate and Desigan's case study, and weigh in with your opinions on the boards below.