Key-Work Consulting GmbH (www.key-work.de
) provides Enterprise Marketing Services for a wide variety of clients primarily in the retail sector in Germany. We have been in business since 1999 and have over 35 employees. We provide hosting services for fully outsourced Business Intelligence to our clients. We perform numerous ongoing data analysis tasks for our clients and also make recommendations which often turn into joint development projects. Our ‹ems
Enterprise Marketing Studio product includes components from Instant Business Intelligence. We manage our own data center with large numbers of computers in our head office in Karlsruhe. Our typical client is a catalog retailer or specialty retailer with 200 mio EUR or so in revenue, and between 300 to 2,000 employees.
We first met Peter Nolan in the late 90s when we went into the BI Business. We met him via online forums. While following the excellent advice he was providing online, we became keen on learning from him, as we were new and he was obviously very experienced. We soon learned that ETL was the core issue in data warehousing. In 2002 Peter told us about a product he was bringing to market for ETL. We had invested heavily in Microsoft’s DTS and so we did not pursue Peter’s ETL product. However, we stayed in touch with him. In 2005 Peter launched IBI and brought out a new version of the ETL product, and it was then known as SeETL. We also declined this version, as we believed Microsoft when they said SSIS would be much better than DTS.
When SSIS finally arrived in late 2005, we still had issues with it. After some time of testing it and learning SSIS, we decided that it was time to take a second look at SeETL. We called Peter in February 2006 to come and explain SeETL in more detail. He gave us the source code to review and was very open about how the product worked. We have people who can write C++, and their verdict was that the software made sense and looked robust. More importantly, we could support it ourselves. We were very taken with the ideas in SeETL, and we thought it could provide us a real edge in a competitive market.
In late 2006 we agreed to a trial of SeETL. We asked Peter to do a piece of work to prove that in skilled hands the tool could do what he claimed. The rest is history for us. Peter showed us that with SeETL, we could very significantly cut the cost of development and support of ETL. We are happy to say that it is significant enough that we have moved all our ETL off DTS and on to SeETL. After evaluating both products thoroughly, we skipped SSIS completely. That does sound amazing, but we can assure you it is true.
Another important thing is this; we thought we were just buying reduced costs and increased productivity. As it happened, we also bought increased reliability! Our ETL systems are far more stable and reliable now than they were before. We have far more control over the schedules than we had previously. It really is amazing how much control we have as compared to our DTS days. This is important to us, because if the data is not there at 7am when our customers come in to their offices, they call us right away, and they are not happy. The number of times this happens has decreased markedly. The overtime hours of nights and weekends has also decreased significantly.
To other companies who build ETL subsystems as a business we have this to say; we have used the C++ engine version of SeETL for more than 5 years now. It is stable, robust and reliable. The workbook provides the simplest possible interface to create ETL subsystems. There is simply no better way to build ETL subsystems that are reliable, stable and robust.
If building ETL subsystems is your business, and you want a competitive advantage, even if you also use another ETL tool in production; we have no hesitation recommending SeETL.