Parking into small berths remains difficult for unskilled drivers. Researchers had proposed different automatic parking systems to solve this problem. The first kind of strategies (called parking trajectory planning) designs a detailed reference trajectory that links the start and ending points of a special parking task and let the vehicle track this reference trajectory so as to park into the berth. The second kind of strategies (called guidance control) just characterizes several regimes of driving actions as well as the important switching points in certain rule style and let the vehicle follows the pre-selected series of actions so as to park into the berth. Parking guidance control is simpler than parking trajectory planning. However, no studies thoroughly validated parking guidance control before. In this paper, a new automatic parking method is presented, which could characterize the desired control actions directly. Then the feasibility is examined carefully. Tests show that a simple parking guidance control strategy can work in most parallel parking tasks, if the available parking berth is not too small. This finding helps to build more concise automatic parking systems that can efficiently guide human drivers.