Deploying a New Enterprise System

Deploying an enterprise system is an expensive proposition, not just in terms of costs (development, implementation, licensing and maintenance), but also in terms of dedicated resources and time. In every domain, there are many vendors promising that the business will instantly run smoother and increase profits if using their systems,  but they fail many times to come up with a viable long-term (or even short-term) road-map.

Here are 10 recommendations to increase the odds of a successful system implementation:

1. Make a clear and extensive list of requirements before start looking for solutions (with focus on specific business processes).
2. Carefully evaluate your options before selecting your system. Better is to find a partner that specializes in your industry.
3. Run a POC (proof of concept) as close as possible with your real business processes.
4. Get references by asking the vendor for at least three happy customers.
5. Consider the amount of customization required to configure and deploy. (A turnkey solution may offer less flexibility but more stability, and less initial and ongoing cost.)
6. Factor in organizational change management. This is pivotal for the project success.
7. Don't forget mobile users.
8. Appoint an internal product champion.
9. Provide the necessary time and resources for training on the new system.
10. Get upper management support. (Despite that is the last point in the list, probably it is the most important one.)

Current software engineering process: 6 topics for improvements

From my experience I see the following areas that need changes:

• Despite using high-end technologies we should not forget that we are (still) working with people. (I should think not at two AWS resources in my team, for example, but at two teammates with the same skills, one of them just having a fight with his girlfriends. Each should be treated in a different way.)
• Sometime software development is done without any communication with the end-user; functionality has more priority than usability.
• Missing business knowledge: with elementary business knowledge it will be easier to develop IT systems to support the business and not to change the business to comply with IT.
• No standardization (for example, simple rules to make code understandable not only by the person who wrote it); missing code reviews.
• Documentation is not kept up to date.
• Testing is not considered from the beginning; software engineers are not doing unit testing.
• No analysis is done after finishing a project, no lessons learned are taken.