What do a Mongolian stir-fry restaurant and a medical lab providing home testing solutions have in common? They are both innovative services that base their success on customers controlling part of the service delivery. These providers allow service tasks to be performed by the customers as a means of shaping the overall experience and not strictly as a means of "outsourcing" the service. Motivated by such practices, we explore whether and how should providers allocate the control of different tasks of their service to the customers. We model services as multi-step processes with each step affecting customers' experience at other steps. At certain steps the provider may hold an “expert" role and be more capable of performing than the customers, whereas at other steps she holds an “administrative" role and is less capable of performing than the customers. We distinguish between routine services, where the service outcome must conform to standardized specifications, and non-routine services, where the value of the service outcome relies on subjective dimensions. We show that the optimal design is determined by an economically intuitive rule whereby the provider controls the steps based on the marginal benefit she can derive compared to self-service. For routine services, this rule translates to managing “blocks" of steps because the provider benefits from containing the volatility of the experiences across the service even when this implies the provision of service steps with a negative marginal benefit, i.e., steps which she is less capable of performing than the customers. Instead, in non-routine services providers should focus on the value advantage they can ensure through a "core provision" even if this implies forgoing control of steps for which they are more capable of performing than the customers and from which they can derive positive marginal benefit. This implies that in non-routine services the provider exercises more control up to a certain process length; beyond that she delegates more steps to the customers. When customers differ in their abilities to perform the different steps, the provider may offer a service line. Service lines facilitate better segmentation than a single service offering, but their economic benefit exhibits an inverted “U-shaped" relationship with respect to the number of steps that a service comprises. Finally, we find that competition between two providers who differ in their capabilities to perform a service results in service design differentiation where the more capable provider offers a higher-end "focused service" against a lower-end "super-service" offered from the less capable provider.