In-Depth
All Shook Up: Open Source BI Disrupts the Status Quo
About one-quarter of shops use open source BI tools; even more are evaluating them.
Just how popular are free or open source software (F/OSS) business intelligence (BI) offerings? Open source vendors like to talk up a looming tidal wave of F/OSS adoption, and there's certainly no shortage of activity in the F/OSS BI and DW markets. What's more, plenty of market forecasts project that F/OSS business intelligence will, over the next half-decade, emerge as a force to be reckoned with.
That change may occur sooner according to a new report published by author and industry veteran Mark Madsen, a principal with consultancy Third Nature.
Madsen's research, which was sponsored by open source software vendors Jaspersoft and Infobright, paints a picture of a teeming open source business intelligence segment that -- although lagging behind the maturity of F/OSS offerings in other spaces -- is quickly evolving, such that it now comprises a credible alternative to proprietary BI. The proof, Madsen says, is in the deployments.
About a quarter of all companies -- from small and midsize shops to large enterprise environments -- are using F/OSS offerings today. (F/OSS use ranges from a high of 32 percent among small shops to 27 percent among large enterprises.) F/OSS is clearly here Madsen points out. What's striking, he continues, is how many shops are evaluating F/OSS BI or DW offerings: nearly 40 percent of small shops and about one-third of midsize and large organizations are mulling F/OSS BI offerings.
The upshot, Madsen suggests, is that F/OSS business intelligence is crossing a threshold of sorts -- moving from niche or tactical to mainstream deployments. That's true of adoption in shops of all sizes: "One persistent myth is that small companies are the primary users of open source," writes Madsen, who stresses that this is no longer the case (assuming that it ever was). "While there are more small organizations evaluating and using open source than mid-sized or large … the data also shows that both small and large organizations are leading adoption over mid-sized organizations."
How shops are using F/OSS tools is also changing. True, small shops are most likely to deploy F/OSS BI offerings on an organization-wide basis. Deployments in mid-sized and large organizations tend to be confined to individual business units, or, smaller still, tactical use-cases, but mid-sized and large shops are expanding their open source vistas, Madsen stresses.
"[S]mall organizations are more likely than medium and large to do company-wide deployments, and large organizations are doing smaller deployments," he concedes, adding that -- from small to large shops, and regardless of the scope of their deployments -- F/OSS adopters tend to share similar characteristics: they're operating with limited budgets, have smaller user bases, and tend to have more "uniform" deployment scenarios.
"Despite this general pattern, there are enterprise-wide deployments of open source in large organizations," Madsen continues. "Forty percent of large organizations plan to or have deployed a BI or DW application corporate-wide with some open source components, demonstrating a level of software maturity."
Similarly, most open source BI applications today support a comparatively small user base: most deployments (54 percent) average between one and 24 users; less than 10 percent support 500 or more. However, the user base that consume F/OSS business intelligence technologies is poised to explode over the next two years. By 2011, Madsen says, almost 20 percent of open source BI deployments will support 500 or more users; by far the largest segment -- at nearly one-third of all deployments -- will support between 51 and 200 users.
Deployment expectations correspond to size, he says: large shops anticipate rolling out open source BI to a large number of (500 or more) users, small- and medium-sized shops have less ambitious -- but by no means reserved -- plans, typically projecting F/OSS deployments that range between 50 and 200 users.
"The number of users in environments with open source is similar to what is reported in actual usage in proprietary data warehouse environments," where shops actually use fewer licenses than they've paid for, Madsen writes. There's a sense, he says, in which the economics of commercial software licensing -- where shops intentionally buy more than they can use and inevitably use less than what they've paid for -- both work to the advantage of F/OSS business intelligence offerings and illustrate one of the primary ways in which open source is disrupting the enterprise software market.
"Open source has an advantage over proprietary solutions regarding scope because there is more deployment flexibility," Madsen continues. "People often buy more software than they need from traditional vendors because the high license cost makes obtaining funds for more seats a challenge, and because of the way the software is discounted in volume purchases. Open source can be less expensive and F/OSS versions carry no penalties for increasing or decreasing usage."