Being a startup is “great” as a feeling. Startup culture is filled with so much positive energy to get the things done. In this process of getting things done, one thing we miss is the proper design in a data platform. It is understandable that people start with a simple data platform and evolve it over the time. Starting with the perfect data platform is less practical when we consider the cost involved and the lack of domain knowledge in initial stages. We should all admit that proper data platform costs a lot, which sometimes not efficient for a startup. My personal opinion is to start small and to evolve with time. Here we will talk about common problems that we faced in a start-up data platform.
Scalability issues impact in several ends. Startup systems are not meant to scale until the end of time. Sometimes they become impossible to scale, sometimes scaling requires so much additional effort that they need a separate team working on scaling the data platform. Sometimes scaling is involved with a large cost that is rapidly increasing. Sometimes scaling increases the overall system complexity and reduce maintainability. If I summarize main impact area of scalability costs, it will be as follows,
- Being impossible to scale
- High Cost of scaling
- Increasing manual tasks of Scaling
- Increase in system complexity while scaling
- Reduction of system maintainability
Proper data platform design should answer above concerns. Proper design should be scalable beyond the foreseeable future. While scaling it should minimize the cost additions, remove any complexity additions and should involve minimal or no manual effort.