Improvements to My Organization:
Our customer is an enterprise that has over 120 entities. It used to take a lot of time (around 15 days) to collect data and generate reports for their performance. After OBIEE was implemented and some processes improved, our customer is now able to review their monthly performance with a clearly analyzed pattern within 5 days.
Use of Solution:
I've been using it for three years. I started with OBIEE 11g (188.8.131.52), and upgraded it to 184.108.40.206 and 220.127.116.11 afterward.
We did not encounter any serious issues when we deployed the tool.
OBIEE is a little unstable in our environment. The server/service has to reboot/restart every couple days, otherwise, users will see errors when they access reports.
I am not going to rate Oracle's support high. It took too many times to collect logs and screen shots, and talk to a support guy. And they still have to go back to R&D once they confirm the issue was product related. It took another loop to communicate with R&D guys, and there almost 70% of thje issues were treated as a "Fix in the Roadmap," which means we won't get a hotfix. We suffer quite a lot when we have to tell our customers that some bugs will not be fixed until it's in the roadmap, and even then, a few years later.
Our customer choose OBIEE as their reporting tool before we assisted them with implementation. I used TIBCO Spotfire after this project, and it's quite a different product. It seems business users will use a so called "analytic tools" such as Spotfire more often than "BI tools" like OBIEE or BO. Since their requirements become much more dynamic, they may not even know what they are looking for, or what questions they suppose to answer. BI tools needs too much IT effort for ETL and modeling to enable users to read the report. However, some questions cannot wait that long, and users need to know the answer as quickly as possible. On the other hand, analytic tools just need raw data and users can those data to generate reports whatever they need for whatever questions/purposes (if they know what those data means well and have some basic knowledge with database).
The setup is quite straightforward, unless you need to deploy with a cluster. Components in a model/presentation layer can do it simply by dragging and dropping, if all the calculations are done in the database. Although, some settings,such as "Write Back", is not that easy. They need to be configured, but you have no idea where to configure it unless you go through the documentation, and even some blogs perhaps.
I helped implement it as part of the vendor-side team.
Just keep the model as simple as possible, it'll make works much easier afterward.
Disclosure: I am a real user, and this review is based on my own experience and opinions.
Aug 13 2015