I think it may be possible if one establishes a grid of stationary countermeasure-launching platforms, creating a mesh above the protected area. The density of the mesh would be a function of the countermeasure acceleration and final velocity figures, in relation to the same of the incoming missile. Further, deviations in the incoming missile's flight path, such as through 'piloting,' would necessitate a tightening of the mesh.
An advancement of this tech would be non-stationary. It could be the function of a more chaotic and transient motion, reactive, and where the minimum density per area is preserved. Moreover, it would be preferable if the mesh itself was the countermeasure, meaning, carrying a payload or of sufficient kinetic energy to destroy the threat. Individual components of the mesh could be governed in simple AI, similar to bird flight. This would allow somewhat decentralized control as figures such the mesh density to be a function of the number of deployed units and autocorrecting. The countermeasure units could be accelerated very slowly and cheaply as this would be done in advance, and would be able to exhibit initial velocities much greater than that of the incoming missile.
Of course, these countermeasures would also be capable of kinetic energy bombardment of the surface.