Comment contrôlez-vous vos coûts de développement de produits ?

Don't use an MVP to validate your idea. If you are building an MVP, that means you have already done your homework and you have come to the conclusion that the product is viable and feasible to implement. A great way to showcase your idea is by using an interactive prototype. You can build one for free or at very low cost using a tool like Figma. Remember it's better to keep it as realistic and close to the final product as possible. It doesn't have to be perfect, but a close representation of a web app with its interactions will allow users and investors understand the product and give you valuable feedback. Do this early in your validation process to avoid costly development. An MVP means you are ready to sell.

This is where the MVP comes into play. It means you have validated your product, your have confirmed there is market for it, and users and investors are willing to participate. Now, instead of jumping into the development of a fully functional and polished product, start by releasing an early version that can start selling. Make a list of the features that the product can't do without and prioritize them. If you are building a mobile app, start with a hybrid approach to save costs. Don't fill it with animations and heavy processing functionality. Remember, make it user friendly and include the core functionality. Now that you have an MVP, you are ready to sell it on the market. Once the product starts selling, you can add more features.

Choosing the right tools can be crucial to your product's success or failure. Research your options well before you decide what to implement. For instance a well implemented EC2 cluster on AWS can quickly run up your bill even before you deploy your application on it. If you are on a limited budget and you are just starting, look for alternatives like Linode or DigitalOcean. If you are building an MVP, evaluate the need of a server and back end. It might be better to reduce the costs from them by using Firebase. It has a pretty generous free tier which can well be used for your initial 5-10k users before you get the need to scale up. With time, as your MRR grows and your product evolves, you can start moving towards more robust tools.

A common mistake startup entrepreneurs make is to spend their initial funding on a sophisticated team, an office and all sorts of fancy things. There's no better way to blow your budget and get nowhere than that. At those early stages, outsource the operations you can't do your self. For example, if you are a techie, you can probably build your Prototype and MVP yourself; if you need help, hire agencies or freelancers. But you are likely not skilled in marketing and sales - partner with an agency specialized in that. If you are good with sales but not so much with tech, hire a professional team to develop your product for you. Once you start selling, then you can start hiring your dream team and transition the work to them.

Tracking your progress is a tricky part of product development and there's no one size fits all when it comes to software development. Metrics is the most inaccurate way to track software development. It's not the same to fix a typo, as to implement a payment system. You can't measure performance by the amount of tickets. On the other hand you can measure milestones. If the desired features and fixes have been delivered in the milestone, then you can mark it as done. If you keep getting the same issue in every second iteration or something has been dragged for several iterations, then you might have a problem with the team. You will have to re-evaluate their capacity to deliver.