What does DevOps have to do with MVP?
The Minimum in MVP has always been about features - just enough features to begin measuring value and sparking additional discussions for future work. The Viable portion often translates to something that can be pushed to production and used by real customers. But, when discussing a new product and what makes the cut for MVP - it can be easy to forget about the glacier of work necessary to deliver those features and measure that value.
Much like deciding which product-features make the cut, the team must identify what platform-features are necessary for MVP release too. In this context, the platform is defined as everything it takes to deliver, measure, and support the product.
The product feature-set tends to take an external approach by gathering requirements and thoughts from customers, domain experts, and market trends. While the platform feature-set takes a mostly internal approach.
Some questions that should arise:
- Is what we built secure? In what ways does it need to be secure?
- Can what we built be supported by us? Who else has to help support it?
- How do we know its working? Is it being used how we expected?
- Are we following standards? best practices? Are we comfortable with how it was built?
Many of these are explored, answered, and accepted by the team itself - which is important for taking ownership of what is delivered. In answering these questions though - the team should engage and reach out to their neighbors: operations, security, customer service, and anyone that can provide guidance on how the product might be delivered, measured, and supported.
At the end of the day its all about setting expectations and being transparent. If what the team defines as an MVP doesn't include performance testing - share that. If it means you there won't be a disaster recovery plan - share that. If the team isn't planning a vulnerability analysis - share that.
Share why you are comfortable with the decisions made around the platform and ensure all your neighbors and stakeholders are just as comfortable. Work together to define what is acceptable to put into a customer-facing environment.
Creating a platform from scratch, even an MVP one, can translate to a lot of work - plan for it! If you are more concerned about testing if your product is even a problem worth investing in to solve, perhaps pivot to building a RAT!