At an online service I attended this week, the minister shared a joke about an argument that took place between three preachers. The three in question debated the best position to be in for prayer---standing, sitting, kneeling or lying prostrate. The joke had a hilarious punchline that emphasized the purpose of the joke--that there is no such thing as a "best" physical position for prayer. The debating preachers were clearly missing the point.
As I chuckled, I thought about how missing the point is also a corporate phenomenon. It brought to mind a data system project that was derailed because some of the leadership team members seemingly missed the point of the project, which was to implement a new ERP system to support manufacturing. The managers each had their own agendas about what the project should achieve and could not agree on the outcomes---or rather they all wanted to achieve their additional desired outcomes.
Project management theory teaches us that the project scope must support the organizational mission. When the organizational mission changes several times over a short period of time, or if there are too many objectives, the result is a project with frequent scope changes. In the case of the project I supported, frequent scope changes meant that project management efforts suffered, and determining a project end date was nearly impossible.
I have never been on a project like that before, and I hope (and pray) that I will never stumble into another. Rather than spend hours lamenting the experience, I have instead reflected on ideas that might have led to greater project success, one of which I will discuss here.
Agile project management uses incremental development to produce a usable product at the end of each delivery phase. With an enterprise system, it is not likely that a usable, yet partially delivered, product would be accepted so usually a hybrid approach to agile development is taken. A proposed technique to incorporate into the hybrid methodology is to stand up semi-permanent system interfaces to allow modular/functional deliverables.
For example, in the case of a third party system, the vendor/supplier management module could be delivered via the third party system while the remainder of the operations (such as the materials management or human resources) would remain on the old system, connected through an interface. This is not unique interfaces are frequently developed to integrate systems on a permanent basis. The recommendation here is that the two systems (old and new) would be administered simultaneously until every function would be converted to the new system. Essentially it would be similar to operating the system in a long-term end-to-end testing environment.
