We are small. We don’t need an SDLC.
Most beginner founders I’ve met had the fantasy of a two steps SDLC process: 1) I have an idea, 2) code my idea.
Then the idea becomes a moving target. The programmer either loses her mind or codes whatever she feels like.
The SDLC (Software Development Life Cycle) is a process to help with this problem.
For MVP-stage startups, it might look like this:
- Estimate and Budget
- MVP Requirements
- Architecture design
- Project plan
- Development
- QA
- Release
The SDLC will take different shapes or models. Don’t get caught thinking too much about which model is the right for you. Pick one, try it out, and adjust.
It is not a one-time decision: new features, customers, and team members will push you to upgrade your SDLC.
I’ve seen startups not planning and spending millions on refactors. And I’ve seen startups spending too much time on product design without shipping anything meaningful.
The goal is not to neglect any steps in the SDLC process and go through them as quickly as possible. It is all a nice story if you can’t ship something to the users.
- Challenges in Implementing Federated Learning in Ad Tech - 10/09/24
- Graph Neural Networks for RTB Systems - 10/02/24
- Measuring Emotional Impact in Local Campaigns - 09/25/24