Nowadays, the optimization of microchannel heat sinks involves using algorithms, e.g., topology optimization (TO), to find the optimal layout of fluid channels to enhance certain objectives, e.g., heat transfer, pressure drop. The outcome is an unintuitive material distribution in the design space, subject to constraints such as manufacturability, fluid flow requirements, and thermal performance. This approach has been shown to provide significant benefits compared to traditional design methods, leading to increased cooling performance. However, depending on the application, defining an initial (i.e., starting) guess for the optimization problem is not trivial. Initial guess for TO, i.e., starting design for the optimization process, can play a crucial role in determining the final optimized design. Its choice may affect the convergence rate, the performance of the final solution, and the computational effort, as it can be set in a variety of ways, e.g., a random distribution, a preconceived design, or a combination of these two. In this study, we propose a heuristic approach for incorporating a genetic algorithm (GA) as outer algorithm into the TO routine in order to find the initial guess that ensures the lowest thermal resistance while limiting the pumping power. The aim is to correlate initial designs with final objectives and to define new useful insights on microchannel optimization, since TO has proved to be a promising area of research for the development of advanced thermal management systems.