Lately, we at Xtivia have seen a marked uptick for interest in open source ETL tools. Each customer situation is unique, but I would like to propose a few factors that you should review before making the decision:
Incumbent Tool Set: Decisions around your choice of an ETL tool are affected by a lot of different things. End of the year or introductory pricing always looks attractive and purchasing decisions are made in the "heat of the moment". This works out okay for a little while but then the costs start adding up. The key is to focus on "how" the tools are being used as opposed to "which" tools.
It is a big issue if you are not leveraging the tool to the fullest extent and your developers then turn to tried and tested method of creating custom scripts/code to facilitate ETL. This further diminishes the value you can extract out of your investment.
Changing Conditions: Change is the only constant, so why would the ETL landscape be any different? Another key factor is to check whether the tool set you are considering works with current and future data sources/destinations without adding to the costs. In other words, would you need to buy a license for Sybase IQ connector if you decide to use the database for data mart purposes?
Support and Community Activity: Open source based tools are just that--open source based. The key is to find the right vendor who can provide enterprise grade support and the right partner who can help with best practices and may be even providing the first line of support. Also, the community that is contributing should be thriving--in other words, your best chance lies with the community that is seeing a lot of activity.
Reliability and Functional Diversity: Key questions to answer are--Once you create a job using the ETL tool, does it function the way you designed it? Is there a "platform" to sustain the ongoing collaborative development? How easy is to tune the jobs for performance?
It may seem overwhelming, but this is precisely what we do at Xtivia. Our tool of choice is Talend. We have worked with many other ETL product suites like Informatica and Infosphere Datastage as well. So, if you are looking to find the right partner, look no further--we can help with all aspects of data integration and specifically using the "best fit" tool for your organization.