We're an online retailer running iStore 3i against an Oracle Apps 11.0.3 installation. iStore was originally implemented because it promised support and services for the web site presentation (iStore), the transaction manager (OAS), the business system (Oracle Apps) and the databases through a single vendor. Oracle consulting was responsible for the installation, most of the maintenance immediately after the go-live date, but was phased out after about half a year.

This was an early installation, and Oracle was very cooperative in customizing the integration and matching them to our needs. However, the iStore did not meet our business model very well. We've had issues with the integration of customer tracking, discount and coupon implementations, account based sales, credit card processing and the scalability of iStore itself.
The scalabiliy issues were mostly the result of iStore trying to cover a lot of ground, thereby including functionality that we would never use.
The customizations also made the implementation very proprietary. We're so far not considering a move to 4i.

Instead, we've phased-out in Oracle iStore 'in pieces', meaning that we still use part of it, but replaced certain functions to better match our needs and increase speed. Hopefully, we'll be able to sever our ties with iStore completely and move from OAS to maybe 9i.
Please don't get me wrong, iStore worked well as a primer for a web store, but it did not meet our business model well enough.
It also has the advantage that it can be easily integrated with Oracle Applications (Java, PL/SQL).

The fast pace of the implementation of the integration also resulted in a lot of bad transactional interfaces that needed to be rewritten. If I were to do this over again, I would force a more defined methodology for transaction processing and try to match the business model closer to the web application.

Recommended: No