After watching agile development for two days, I summarized the essence, what is MVP?
MVP is not a poor product release, it is the smallest product release that can produce the expected results; MVP is the smallest experiment to verify the hypothesis, and the version iteration of the product is the result of continuous experimentation until the product is proved to be correct.
-
MVP is validation-based learning. Proceeding from hypothesis to verify, hypothesis-driven development model is one of the core concepts of Lean Entrepreneurship. To know what problem to solve, to understand that the current solution is only a hypothesis, and even the problem of solving h is just a hypothesis. Starting from understanding the hypothesis and quickly verifying it, each step and each function and release has a clear goal, that is study. This is what Eric Ries calls the "develop-measure-know" cycle. When you observe that users have started to use the product intently and recommend the product to others, you know that you have achieved the minimum and feasible. At this time, it is time to bring the product to the market. If it is launched before then, the result is bound to bring a lot of disappointment. customer of.
-
Products are discovered gradually, and demands are gradually excavated. The process of product production is more like the process of a baby’s birth and growth, rather than being able to run and jump after birth. If the early development is already a complete product , It can only be said that there are too many functions developed.
The user story map is a good tool for exploring MVP: through the user story map, you can divide the beginning, middle and end of the product.
-
Start: Focus on must-have features and focus on technical challenges or risks. Skip the steps outside the main process and develop the main process regardless of the business rules that complicate the problem.
-
Middle game: Supplement the peripheral functions and start testing the non-functional requirements of the product.
-
Finale: polished release, more eye-catching and more efficient.
MVP is to provide a complete experience, not an incremental model, not just develop some modules, but need to run through a complete minimal business closed-loop process. As shown below. We are more used to the analogy of cutting a cake. MVP is a top-down slice of a cake, not a layer of it.
The process of MVP and Lean Entrepreneurship is a process of scientific trial and error, a process of finding a way instead of running; it does not help us grow faster, but helps us reduce stop and waste time, make more accurate steering, and adjust faster; Lean entrepreneurship emphasizes proven cognition and needs to rebuild the concept of learning: there is expectation, verification, and cognition; the MVP small batch method allows new companies to minimize the time, money, and energy that may be wasted; reduce Batches, complete the develop-measure-recognize feedback loop faster than competitors, and the ability to understand customers faster are important competitive advantages that new companies must have.
-
MVP is aimed at validating basic business assumptions, and selects the smallest entry point on users and products.
-
MVP is only aimed at early angel users. This group of people has a higher tolerance for the product, can see the future of the product, and is willing to interact and improve the product together.
-
Only serve the smallest range of users, and ignore other things that can be ignored.
-
Only keep the most core functions, and don't want anything else. In terms of product functions, it is recommended to cut the imaginary product in half, and then cut it in half again, so as to achieve the real minimum function combination.
-
Functions must be subtracted, and only the lowest scope of development should be done, and those that can be pretended will not be done first. Just like the demo video of Dropbox, MVP can be achieved without a single line of code development;
Based on the above user and functional assumptions, design measurement and data collection strategies, collect quantitative and qualitative data at the same time, conduct effect analysis through methods such as A/B testing, cohort analysis, and net promoter score to avoid vanity indicators.