When it comes to data modelling there are two camps at the forefront. Those who consider it a skill and those who consider it a dedicated role. In this debate there is no real winners since the answer lies firmly in the long grass of it all depends. Let me illustrate the point with a story.
A few years ago, my father built a shed out the back garden. With dogged determination he spent weeks laying down brick, mixing cement, and sawing planks of wood in half. To everyone’s surprise he managed to complete it without seriously harming himself. Even more surprising was the resulting structure stood and still stands today, albeit with crooked walls and a leaky roof. Not bad for a man with no building experience and tools that consisted of a measuring tape and rusty saw.
The point of this long-winded narrative is my father knew next to nothing about architecture but had enough to get the job over the line. Believe it or not database projects in the real world are often run in a similar manner. Data modelling is rarely considered a full-time role, instead this responsibility is often dished out to existing personnel. Usually application developers. For small to medium sized projects this is not a problem. Most developers will have data modelling in their tool-belt and will have enough knowledge to do a competent job. Any inconsistencies and inefficiencies at this scale can be easily papered over.
Things get messy, however, when the project grows in complexity. Just like my father’s shed-building ability doesn’t equip him with the skills needed to tackle a skyscraper, a developer might not have a sufficient handle on things to take on an enterprise-sized data modelling job. For large projects security, storage, and reporting all need to be taken into consideration, and this might be a bridge too far for developers. It is at this critical juncture data modelling crosses the threshold of being a useful skill to becoming a full-time role.
Data modelling, for those who are unaware, is the process of designing, building, and documenting the database architecture of a company. It involves mapping objects and behaviors to relevant database counterparts. This is a hugely important step in the development process. Data modelers are essentially the bridge between the business requirements and the technical implementation and as such need to be fluent in multiple database systems as well as being able to interpret business functions at a minute level. Being able to walk this tightrope between the two sides is a rare breed in the IT world.
In recent years the job has become more and more relevant. This is due to the rise of data science and machine learning. Across the globe businesses are becoming increasingly reliant on data for their decision making and making predictions about the future. This had led for need for clean and efficient data that is optimized for reporting and analysis. Expect to see Data Modelling job vacancies over the coming years.