Cost has, more often than not, been cited as the primary reason behind the reluctance of CIOs to opt in for traditional Business Intelligence (BI). Though BI is an established buzzword in the Indian market today, there are no doubts around the fact that it comes at a pretty substantial cost.
It is precisely this excitement around BI, coupled with the cost factor that is clinching the case for open source BI. While enterprises are keen to experiment with BI, they are not so eager to spend massively on something that they are not yet sure of. As a result, open source BI is beginning to gain a foothold
in India.
According to Nishchal Khorana, Head – Consulting, ICT Practice, Frost & Sullivan, South Asia & Middle East, “With the increasing adoption of business applications by the Indian SMB sector, the need to draw out intelligence from the rapidly multiplying quantum of data is only becoming more critical. With price being the key constraint in the adoption of BI tools by SMBs, open source solutions offer a viable alternative.”
As of now, the ratio between enterprise and open source BI stands at 80:20, worldwide. It is estimated to be somewhere around 90:10 for India. However, while open source BI is estimated to be growing in single digits globally, there is nothing less than a 25% growth that is being estimated for it in India.
The community edition downloads of Pentaho for India alone stood at over 5,000 in 2011, compared to around 3,000 for the preceding two years combined (2009 and 2010). The key industries involved in the downloads were about 65% in IT services, around 20% in manufacturing and retail, 5% in telecom and 10% from others. The numbers are indicative of the interest that open source BI solutions managed to elicit in 2011 and these numbers don’t include the numbers for the other big open source BI vendor, JasperSoft.
Jaspersoft claimed that it managed to grow at 60% in India during 2011. Though these growth rates might be on a significantly smaller base, they illustrate beyond doubt that traction is indeed there for open source BI solutions in India.
Though cost is the first and the biggest driver behind the adoption of open source in a price sensitive country like ours, the maturity of the open source community also has a role to play here. There are, however, some key dynamics that need to be examined.
Sporadic adoption
Despite the fact that open source comes with obvious cost and scalability advantages, it hasn’t penetrated into the Indian fabric, yet. While reluctance indeed happens to be the biggest reason behind limited adoption, experts said that there were some ground realities that had to be taken into consideration.
There are apparent questions around support in open source that are prevalent across the spectrum. Although there is a huge community behind open source BI offerings and the answers to almost all questions are available online or with the community, the perception is that prompt support for these solutions is lacking.
According to Sameer Dixit, Delivery Manager, BI & Analytics Solutions, Persistent Systems, “Organizations often don’t use open source BI for mission critical apps because of support issues.”
There is another opinion floating around with respect to support issues for open source BI tools. Quite a few analysts believed that support issues were redundant in any kind of BI since the role of system integrators is engraved into the BI lifecycle. Bhavish Sood, Research Director, Gartner, commented, “This is no longer a big issue for open source because organizations are mostly dependent on system integrators for support.”
He added that, although support was almost always provided by the SI, some programs, functions or modules might not be as well supported. “Not all community BI solutions are deployable since they require strong SI support and backing.”
Dixit observed, “Since open source BI is a user community run thing, it often lacks in vision unless there is a big sponsor at the back-end.”
Support could also be the major reason behind a lot of enterprises opting for low cost business intelligence solutions rather than open source. Low cost BI vendors like MicroStrategy and QlikTech are giving stiff competition to open source BI vendors. Dixit concurred, “A relaxed licensing model, even from mainstream vendors, is eating into the open source BI pie. Also, there is a lot of in-house BI development that is being done using open platforms.”
“Visualization led data discovery analytical tools are also impacting open source BI because they have a much faster go to market,” added Sood.
A lot of in-house BI, that is still popular in India, is happening on open platforms and this could be an untapped opportunity for open source BI. However, what is happening with open source in India today is much like a market recce before a shopping spree. Enterprises are flocking to open source to get a “hang of business intelligence,” as Nitish K Mishra, the IT manager at a small accounting firm, Xcel Accountants, put it. Mishra explained that though there was only a workforce of 25 people at Xcel, the volume of data that they dealt with was huge. Therefore, Mishra downloaded the community version of Pentaho and started doing some basic reporting using it. It has only been a couple of months since Mishra flagged off his “experiment” and he plans to move to the commercial side as soon as he is sure that BI is yielding value for his firm.
Mishra’s example brings us to a larger question in the open source BI domain namely the dilemma of choice between commercial and community.
Commercial vs. community
It is the community edition of BI software that truly comes with the free tag of open source and is free to download and use. There is also, on the other hand, a commercial open source edition which has a small annual fee for support. This support fee ensures that customers receive timely assistance and support not just from the parent company providing the open source software but also from its regional partners operating in specific geographies.
Lokesh Narasimha, Director, Operations and Co-founder of GrayMatter Software Services, Pentaho’s partners in India, threw some light on the matter. “While the community edition of the BI software is being used successfully by many organizations worldwide and there is a vibrant community or forum to get queries answered, there is no guarantee that a timely response will be received.”
With a commercial open source solution there is the confidence that a SI or a partner brings to the table. So, depending on the criticality of the BI deployment and availability of skilled IT resources in house, organizations can choose the right versions of the software.
“Many organizations are willing to pay support fees in order to get timely support and assistance,” said Narasimha.
Thanks to this, commercial versions of open source BI seem to be doing much better in India. A lot of BI projects begin as experiments by downloading the free software and evolved into commercial usage.
The other argument tilting the deal in favor of commercial BI is that community BI tools are not as evolved as their commercial counterparts. Because of its relative newness, open source BI doesn’t yet cover the range of features and capabilities on the same scale as the older, proprietary BI tools.
However, one of the greatest benefits of going with open source will almost always be the cost-savings, because the total cost of ownership of an open source solution will never be at par with the big software installations. That doesn’t imply however, that there would be no cost at all.
There’s a cost
That zero cost is an implied consequence of using open source is one of the most common misconceptions in the BI space. Although the community version is free to download and use, there are other costs that are associated with BI. As Sood of Gartner mentioned, “While the overall cost of build can reduce drastically with open source BI because of licensing, open source isn’t zero cost. The cost of a BI solution is not just the licensing cost. There are several other costs like hardware, compression, storage etc.”
“The crucial driver for the adoption of Open Source has been its cost-effectiveness from a TCO perspective, the try before you buy approach and faster deployment as compared to a proprietary solution’s turnaround time. However, you cannot ignore the hidden costs or ongoing maintenance while calculating the savings done on licenses,” said Sesh Rangarajan, CTO – BIM, Capgemini Consulting.
The commercial version of an open source solution, that is more commonly used across the chain, does come with a price tag and there are a lot of organizations across the world that are making open source BI available commercially with added support and enhanced features. Jaspersoft, Pentaho and Actuate are examples of such open source BI vendors.
Brian Gentile, CEO, Jaspersoft, said, “Many still think that open source equals free software, but that isn’t necessarily the case. Many commercial companies have formed around the popular open source projects and products, offering more feature-rich or capable versions of these products, only with a commercial license. The most common commercial approach has been termed ‘open core’, where the software provider builds commercial features on top of the open source core, making a new version available through a commercial software subscription.”
That these commercial subscriptions are still generally much less expensive than proprietary competitors and can provide the best of both worlds is what is making the open source proposition so viable.
Open source software is distributed along a pyramid of possible licenses, ranging from totally free software to commercial open source. Some licenses allow for almost any type of use with no or minimal attribution or reciprocity. Other license types are quite restrictive. According to Gentile, users should be aware of the license that accompanies the open source software that they intend to use, ensuring that they comply with the terms of the license and that the license suits their requirements.
Despite the fact that open source doesn’t necessarily spell ‘Free’, its star attraction is its low cost. Narasimha of GrayMatter gave an example using Gartner figures. “Pentaho is likely to cost only as much as 18% of the cost as compared to mega-vendors like SAP and IBM, and 23% of the entire cost when compared to BI solutions from MicroStrategy or QlikTech.”
For years, the proven and promoted selling point for open source software has been the ability to save money, plain and simple. Although costs are important, awareness about the other benefits that open source brings to the table is also on the rise. One of these key benefits is the flexibility that open source delivers in order to evolve with customer requirements.
Who’s it aimed at?
The lack of the right skill sets is perhaps the most commonly given reason behind the limited deployment of open source BI solutions. Sood explained with a contemporary example, “The crucial challenge with open source software is the availability of companies with the right resource pool who can deploy such tools. Google for instance has an army of geeks who can manage their open source BI. Most enterprises don’t have such resources at their disposal.”
Dixit added, “The hurdle with open source BI is that there is hardly any coolness factor attached to learning how to support or deploy open source BI as against commercial options, so the lack of resources is apparent in the market. Not too many people are vying for it, so getting the right skill sets is a bit of a challenge for enterprises.”
“The other thing is that most of the open source tools are not self-serve tools unlike a BusinessObjects or a Cognos where a user can do his own reporting and dashboarding. In the case of open source solutions, you need an IT and development team in place who can keep on raising requests for various things. Therefore, the cost of sustenance and management goes up,” explained Joseph Clarton, a tech blogger in US who writes extensively on BI.
Ironically, despite the resource issues, it is SMBs that are displaying greater affability to open source BI. Although this could have a lot to do with the price points, there are other reasons behind the SMB-open source equation. Observed Dixit, “Guys who don’t have any BI background go for open source because they want to start low cost. Open source BI is being used where a little bit of delay is not going to change the world and that is mostly the case with SMBs.” Training is another big benefit for SMBs. Vendors like Pentaho and Jaspersoft provide reasonably priced training modules and hosted services as a part of their offering.
Also, for a large enterprise investing in open source BI, it may get a faster RoI, thanks to scale, but for an SMB, these costs are huge. The fact that open source is going very niche with small communities building tools over Jaspersoft or Pentaho etc. for micro-verticals is also aiding adoption at mid-sized companies.
As far as the enterprise play goes, it is not the cost but rather the features that decide the outcome of a deal. For instance, the mobility story in open source BI is not as evolved as it is in the case of proprietary solutions. This could well affect an enterprise purchase decision. Since some proprietary BI offerings are packaged in a glitzy manner, with offers and lenient licensing, going with traditional BI has been the choice of Indian enterprises, so far.
Gentile of Jaspersoft disagreed. He said that large enterprises were adopting open source BI and that there was a strong case for doing so. “In the business intelligence landscape, enterprise software providers are viewed as having more robust features in their product set. However, these software providers have built out their BI capabilities through acquisition, assembling legacy software from a different age of technology. Though powerful, the architecture of these legacy systems can force customers to free up lots of server space, incur expensive training and upgrade costs and undergo lengthy transitions when a migration becomes necessary.”
BI builders of the future
As of now, open source tools such as Wicker, Apache and Hadoop are being used mostly around predictive analysis and big data. Additionally, open source BI is frequently used in embedded software projects, where a software developer would rather use an existing BI platform (reporting and analytics) rather than try to create one from scratch. Especially in these embedded scenarios, modern, standards-based, open source code has proven to have a tremendous advantage.
“A de facto open source BI software stack is emerging including open source database, data integration, data quality, data management and business intelligence tools. The latest one is Apache Hadoop that offers open source software for reliable, scalable and distributed computing of large data sets, which has taken over the Big Data phenomenon,” said Rangarajan of Capgemini Consulting.
The fact remains that a substantial chunk of BI in India is happening through the medium of Excel sheets. That might be an impediment, but there is, according to analysts, a lot of low-lying fruit for open source players in that quarter.
One thing that is going in favor of open source is the backing of the developer community. Jaspersoft has a 16,000 strong developer community in India and it is adding about 400 people to that number every month. Pentaho also has a strong developer community in India.
With experimentation in open source BI gathering momentum and support becoming more readily available, analysts predict that open source BI will see wider adoption in India. The fact that the Indian government is beginning to open up to open source is also aiding the cause of open source BI.
Though open source is never expected to match traditional BI in terms of functionality, the industry is of the view that, with support issues being addressed, some reluctance to adopt with respect to open source will fade.
Till then, it will only be about saving costs and not functionality.
