Serval is a recently developed network architecture that presents a model of service-centric networking, decoupling logical access to a network service from the underlying hardware that runs it. I present an incremental deployment strategy for Serval based around the idea of a Serval translator, a middlebox that allows non-Serval clients to communicate seamlessly with Serval services. Using this deployment strategy, service owners can achieve many of the benefits of Serval without the expensive costs normally associated with deploying a new network architecture. A prototype translator implementation demonstrates that the strategy works and is efficient enough to be used in real networks today.