Migration of platforms after realising "mistakes"

The cost of deploying a system is much less than replacing a system not delivering on its promise. The migration of systems from old versions to newer versions are always an expensive exercise. Even just stopping a roll-out process to consider a decision and review alternatives can be costly. When organisations realise that they have made a "mistake" it is even sometimes to late to convert to better solutions.

In a relatively new industry like mobile banking, where clear leaders have not yet emerged and requirements are still dynamic, the selection of a technology solution will often be re-visited. During the past year, this re-evaluation of platform decisions occurred in many of the established deployments. A lot of effort was spent on evaluating alternatives (even after lengthy deployment projects have completed). Many banks have realised "mistakes" and re-tooled (requiring major investments and leading to very long delays in getting services ready). I am sure we are all aware of many such cases.

The mobile banking industry would have been much more advanced if it was not necessary to throw out systems to be replaced by better solutions. This would have been (and can) be eliminated if companies choose platforms with robust architecture based on a proven track record.