Share what you know with millions of people
Focus is the best place to turn what you know into remarkable content
0
Is Open Source BI a viable option? What are the benefits?
We are a growing company that needs to develop more of analytics core competency to support our business. So in trying to move beyond spreadsheets we’ve been looking at different BI and reporting tools. In addition to the usual players, we’ve been looking at companies like Jasper and Pentaho. Are there really any advantage or disadvantages to going with open source BI?
Events
- Dos and Don'ts of Small Business Marketing May 29 @ 11 am PT
- Lead Nurturing 202: The Next Generation May 31 @ 11 am PT
- The Tricks to Paid Media June 6 @ 11 am PT
- Display Advertising for Brand Awareness June 20 @ 11 am PT





10 Answers
Business Objects spent over $260 million on R&D in 2007, when SAP acquired the company for 4 billion Euros. I don't think anyone believes that Open Source BI will overtake the functionality depth and breadth of such industry giants anytime soon. Still, the economics of the Open Source model are sound and will continue to evolve new and better solutions.
OSBI guru Kevin Goodman has scaled to 1 billion row daily - that's daily - delta volumes on Pentaho for Amazon's OSBI cloud initiative. So, you can't say that scale alone prohibits OSBI. (See http://bit.ly/b33usZ for the evaluation which used the http://tpc.org OLAP benchmarks).
The two considerations I see are 1) Is the DW heavily interdependent with the OLTP systems (SAP BW to SAP ECC for example) and 2) Does the corporate IT culture nurture creative, talented developers, who can bring success to the much less rigid world of Open Source?
Many IT cultures haven't changed much since the days of Hollerith Cards. Driven to protect against disruption and block risks, old-line IT managements will stifle Open Source before it has a chance to draw its first breath. But on the other hand, well, there are companies like Google.
For established, conservative corporations that have high switching costs due to embedded OLTP/OLAP hooks, OSBI is unlikely to displace licensed BI. Don't be the champion there.
But for nimble IT organizations -- large or small -- with the development talent, OSBI is an attractive option, in flexibility, in costs, and in raw potential.
Before tackling the open source conundrum, I'd look closely at any incumbent BI and/or analytics solutions, to see if they could be leveraged further at no cost or with preferential pricing from incumbent vendors. And I would not be afraid to use open source solutions as negotiating tools with incumbent or candidate vendors of alternative solutions. And if CRM or related solutions from vendors such as Microsoft, Oracle, Salesforce.com and SugarCRM were already in place or under consideration, I'd give a long, hard look to those vendors and their respective partner ecosystems for useful BI and analytics tools and add-ons.
With all due respect for Doug Grant, I think there are a lot of questions I'd like to see answered way before getting to issues such as data warehouse alternatives. Especially for smaller and growing companies, there's a lot of BI- and analytics-related ground to be covered beforehand, starting with an assessment of what's already in place that might be useful. When looking to add to the available tool kit, open source solutions deserve at least as much serious consideration as traditionally licensed (and priced) alternatives, as Doug rightly points out.
Great question, Adam -- keep the Focus community apprised of your efforts and their results, please!
Completely agree with Michael above. Small and medium companies should have SAAS BI as their first option. SAAS solutions are very affordable and don't involve software (licensing, support), hardware (even open source software need to be installed in-premises) & maintenance costs.
I work for Zoho Reports which has a free edition that supports 100,000 rows and our pricing plans start at $15 / month (250,000 rows). Do have a look at Zoho Reports (http://reports.zoho.com).
Personally, I would avoid open source like the plague. They're not going to save you much on licensing vs. the Microsoft offering. The MS licenses are included with your SQL server licenses, so this may mean free for your organization.
With the MS platform, you get integration with the office suite of tools like Excel, and that's really powerful. The whole reason you're getting into BI is to mine your data and share it across the organization. Open source can lead to proliferation as different people or departments try on the new cool tool. You end up with a really fractured data environment.
Also, in a small organization, you've got to plan for total victory. You are either going to grow or get bought. You're more attractive if you go with a standard enterprise solution, and you will also find it much easier to grow without rearchitecting everything. If you're going to shrink or stagnate, it doesn't matter what solution you choose.
In short, MS can be a low to zero investment solution, and the capabilities you build around those tools will be valuable to your organization and the employees that develop them.
Interesting discussion. Let me join the debate, I'm working in Engineering, the Italian company (www.eng.it) developing and supporting SpagoBI, the completely free open source BI Suite, no distinction between community and enterprise version (see www.spagobi.org), probably less known than Pentaho or JasperSoft in the US but well diffused in Italy and Europe. We strongly believe that the true advantage of OSBI is not only the cost but more the true chance to build solutions that are really tailored to user requirements. The business model we believe in and we are proposing is that of a "project-centric" solution where the project is more important than the product itself and before starting a project you may ask yourself "is OSBI ready to support what functionalities I'm really looking for ? "; the advantage is that if you answer "yes" you can start prototyping without huge investments and try to set up a sort of Return Before Investment, if the answer is "no" you know that you can rely on community and, in the case of SpagoBI, on the editor support to extend the solution. That's the way how we add engines to the SpagoBI Suite like the Geo Engine, displaying indicators on graphical maps provided by GIS or stored in SVG, or the Query by Example to freely query the Dabatabase and many other. In few words the roadmap is truly driven by project and customer needs
One of the best option is BIRT or Actuate
infomation technology
we can break down BI into ETL DataWarehousing Reports Analytics
My experience has been that in ETL and DW OSBI have matured and have an excellent role to play. actually in a startup project we used Pentaho's ETL [KETTL module] to even manage data between services which were required in different forms
When it comes to visual complexity, interface and usability and when it comes to analytics, OSBI have some catching up to do, they would i guess sooner than later.
The enterprise needs to take a strategic investment approach for this problem split the data integration & BI needs into buckets and experiment on various platforms based on the economics. Closely measure the outcomes and then followup with standardization.
Pentaho, BIRT, Jasper has been our experience, though i have heard a good lot about spago
There needs to be a clarification of what open source really means. We have been "selling" Jedox Palo by wrapping it in services. Technically it is open-source as is Pentaho and others. Yes, you can download the software from some techie site and integrate bit in your own applications or develop it yourself but most real people don't want to do that. These "open-source" vendors have used the banner to release free software with some form of limitation. If you want to upgrade functionality, usage the you have to pay a fee for the proprietory stuff that they have put on top of the "open source" database.
As a marketing tool Open-source is great but don't forget that the developing company has to make some money from somewhere in order to keep developing it. PALO is being constantly developed and is becoming as robust as any other OLAP db on the market. Jedox have a very believable roadmap for future development including a very exciting GPU project.
If you ar elooking for further guidance or wan tot ake advantage of our training and implementations ervices please get in toouch
Forrester released a report this week on the Open Source BI, comparing the various vendors, that you may be interested in reading (http://bit.ly/cDi4BB). Actuate BIRT comes out on top.
Answer This Question