A proportional controller makes the control action proportional to the current error:

$$u(t)=K_P e(t).$$

Its transfer function is simply

$$C(s)=K_P.$$

Increasing $K_P$ makes the controller react more strongly to a given error. In many plants this can speed up the response and reduce steady-state error.

Proportional gain also moves the closed-loop poles through the characteristic equation

$$1+K_PG(s)=0.$$

A larger gain is therefore not automatically better. It can increase overshoot, actuator demand and sensitivity to measurement noise, and may eventually destabilize the loop.

Proportional control is the simplest feedback law: it acts on present error only. Its limitations motivate integral action for persistent error and derivative action for anticipatory damping.