Volume ramp ups are notoriously difficult in digital services, where market pressures can lead to ramping up too soon and too rapidly which in turn can lead to the need to ramp down. This paper addresses the challenge of taking innovation to scale in an established firm by enhancing our understanding of the nature of service ramp ups and ramp downs. Digital service ramp ups differ substantially from production ramp ups as the speed is much greater, and problems are visible to customers. However there are similarities between service ramp downs and product recalls and an important contribution is exploring the nature of ramp downs their processes and possible causes. Using an engaged research approach, longitudinal data from three consecutive ramp ups in a European telecom operator were collected. Through analyses of cases, qualitative and quantitative case data, and using a system dynamics model, we identified a set of issues that affect service ramp ups and ramp downs. These include the need to ramp up the service supply chain, biases leading to unrealistic assumptions about scalability and problem‐solving, decision biases in various functions, launching digital services in beta form, a lack of transparency of capacity and lack of learning from previous ramp ups. We show that if these problems are not addressed or resolution is delayed, this can lead to cycles of delay, backlogs and productivity problems and the inevitability of a ramp down. We explore reasons and importance for such delays that lead to service ramp downs.