In this post, I want to talk about motivations for doing data-science. The reasons for doing data-science might seem obvious, but actual practice shows, that this topic is neglected.
In today’s business software we distinguish between On-Premise Software and Cloud Software. In On-Premise Software, the software artifacts are delivered to the customer and is owned by the customer. The customer is responsible for running the software in an appropriate infrastructure (often in the own data-center) and maintaining it. The software vendor has neither influence on how the customer runs the software nor access to the data the software produces and consumes.
In On-Cloud Software, the infrastructure for running the software artifacts is delivered by the software vendor. Compared to On-Premise software, the customer is no longer responsible to run the software or maintain it. This responsibility is handed over to the software vendor.
Obviously, both types have their benefits and disadvantages. For the first time in IT history, (big) companies are moving more and more to cloud software. For the first time, software vendors have the possibility to collect and use telemetry data, to improve their services/products and provide more customer centricity.
In my opinion, the most important motivation for data-science is to provide better services and products. Decisions are no longer based on experience or gut feeling. It is possible to provide some scientific evidence for each decision you make in software development.