With the rapid development of science and technology, big data has been applied in many fields and has brought commercial revolution[1]. The scientific community generally regards big data as "massive data + complex types of data". Commercial applications are more concerned about big data as an analytical (prediction) method and focus on the potential commercialization of analysis results. All walks of life will produce large amounts of data every day. The transition of data-scale brings huge commercial value, which will certainly bring the innovation of business model[2]. Particularly in the internet and other emerging industries, because they get data more convenient and fast. Like Amason, Facebook, Google etc, they use analysis of big data to innovate their business model for maximizing their profits[3], actually business model refers to "an enterprise’s profitable operation mode plus ways to make money"[4]. So the effectiveness of business model innovation of big data on emerging industries has been remarkable. But the impact of big data on traditional industries is still in the exploratory stage. Traditional industry mainly refers to the labor intensive, manufacturing oriented industries, including the traditional commerce and service industry[5]. Learning from the experience of big data on business model innovation of emerging industries, traditional industries can use big data to subvert the business model and accelerate the transformation and upgrading.

Many software enterprises prefer to use usual entrepreneurial approaches and make detailed business plans, which cost them a lot of time and resources. In the highly uncertain environment, the above practice lead to fail at business or entrepreneurship. To avoid the undesirable feature, software enterprises should adopt lean start-up method. The core idea of lean start-up method is to put in a minimalist prototype on the market, and fast iterative product on the basis of customer feedback (Ries, 2011). Because customers are introduced to the products or service iterative innovation process, the enterprise can meet the needs of the market better and grow up rapidly. Meanwhile, the enterprise can find a suitable business model to achieve efficient operation or reduce the risk of entrepreneurship (Maurya, 2013). Combining case study and theoretical deduction, this paper will show a process for software enterprise using lean start-up method. Firstly, based on literature reviews, the paper will sum up the basic logic of lean start-up and specifically expatiate the five-step model for software enterprises which is “idea generation-customer validation-prototype creation-customer feedback-product iteration”. Secondly, this paper will give a lean start-up process for software enterprises. In the process, the first step is drawing a business model canvas rather than a detailed business plan, it may include nine blocks which are key partners, key activities, key resources, value propositions, customer relationships, channels, customer segments, cost structure and revenue streams, each block has a series of hypotheses (Blank, 2013). The second step is adopting the customer development method to test these hypotheses and adjust irrational hypotheses to satisfy customers’ requirements, this step is referred to as validated learning, with the aim of setting up a continuous customer feedback channel and achieving long-term sustainable development. The third step is adopting agile development to build minimum viable product and iterative the product according to customer feedback quickly, in order to reduce waste and save resources. It is essential for lean start-up to use a smaller, faster iterative process to verify the idea (Ries, 2011). Thirdly, this paper will use several typical cases of software enterprises to verify this lean start-up process. Finally, put forward the general rule of lean management so that it can provide reference for the software enterprises.