Managing a startup using methods proven to work in established enterprises does not typically work as expected, as the startup operates under a totally different environment as a large enterprise. While enterprises create and develop products based on a known business model, a startup operates under extreme uncertainties and changing conditions. These different conditions have led to the development of new tools and practices for startups with little resources to develop new products. This thesis investigates the use of the Lean Startup method using two different startups, Movendos and Taplia, as case studies. It is said that Lean Startup can be used in startup environments where the company is working under uncertain conditions to quickly learn about its business space. The methodology focuses on fast iteration of product and feature concepts based on minimum viable products. This quick iteration and delivery of product to clients produces immediate feedback from the clients and allows the startup to develop their business practices. The case study showed that the two companies had implemented a slightly different subset, but both utilized the MVP concepts. It was found that the lean startup concepts seemed to allow the companies to quickly iterate their products towards a viable one. However both companies had problems analyzing when a minimum viable product was actually viable. Testing a full product with a subset of features sometimes led to situations where the MVP did not really emulate the full product. Another problem was the resource intensiveness of the testing and iteration. Both companies had to make compromises and not use all the recommended tools in the Lean Startup method. The Lean Startup seems to be a decent framework for setting up a software business as its iterative nature fits well with the agile software development. Both companies under investigation reported that they had benefited from the lean startup ideas. As such the lean startup method should not be taken as total method for developing software, but as a toolkit from which to pick tools that suit your software product.