It is well established that porosity in carbon materials can benefit electromagnetic wave absorption by providing stronger interfacial polarization, better impedance matching, multiple reflections, and lower density, but an in-depth assessment is still lacking on this issue. The random network model describes the dielectric behavior of a conduction-loss absorber–matrix mixture with two parameters related to the volume fraction and conductivity, respectively. In this work, the porosity in carbon materials was tuned by a simple, green, and low-cost Pechini method, and the mechanism of how porosity affects EM wave absorption was investigated quantitatively based on the model. It was discovered that porosity was crucial for the formation of a random network, and a higher specific pore volume led to a larger volume fraction parameter and a lower conductivity parameter. Guided by the high throughput parameter sweeping based on the model, the Pechini-derived porous carbon could achieve an effective absorption bandwidth of 6.2 GHz at 2.2 mm. This study further verifies the random network model, unveiling the implication and influencing factors of the parameters, and opens a new path to optimize the electromagnetic wave absorption performance of conduction-loss materials.