Data-based solutions powered by artificial intelligence (AI), and especially its subdomain of machine learning, are a key driver of today’s fast-paced technological evolution. In the process industry, barriers for many organizations to apply this technology are missing know-how for conceptual planning as well as lack of economic feasibility studies. However, companies risk to lose their competitive position by not applying this technology. In this article, we describe the CRISP-DM model as a conceptual planning approach. In addition, we provide practical advice based on experience in other industries how this technology can be applied in the process industry. Here, the steps process analysis and data understanding are key success factors in order to develop economically viable use cases. An implementation strategy should include an agile environment to develop ideas fast and with little risk before transferring working solutions to the requirements of the operational business. A combined bottom-up / top-down approach of knowledge distribution and pilot projects can help organizations to successfully embrace this technology in their operational businesses, overcome associated fears and organically seize company-individual opportunities that arise.