Although Earth's surface motion is well known, the flow field in the underlying mantle is not. Mantle flow is typically calculated on the basis of inferred density variations, and flow directions can also be reflected in seismically observed anisotropy, but those observations leave ambiguity on the depth and direction of the deformation. Anisotropy orientations in East Asia, outside Tibet, have been interpreted in various ways and have often been linked to deformation in the asthenosphere related to absolute plate motion and/or mantle wedge deformation. Here, we re-analyze published seismic anisotropy data and find that orientations outside Tibet can be much better explained when considering absolute plate motion (APM) of the Earth's surface in addition to coherent sub-asthenospheric mantle flow, than when comparing orientations to APM alone. The direction and magnitude of the required sub-asthenospheric flow depend on the absolute reference frame used for the surface velocities, but when considering an APM frame with an intermediate global net-rotation we find an eastward flow of 1-2 cm/yr. This flow is faster than the surface motion, and generally in the same direction, from which we conclude that the mantle leads the plate motion. Our inferred flow is similar to those independently calculated based on buoyancy forces driven by density variations, most notably the high density anomalies associated with the western Pacific subduction zones, but possibly also the upwelling underneath Africa. Additionally, based on our predicted sub-asthenospheric flow and absolute motion of the lithosphere, we predict asthenospheric-based XKS orientations underneath all of east Asia and find it to differ significantly with observed XKS orientations where either the lithosphere is thick and/or strain rate is high, which suggests that at those places observed XKS orientations reflect the integrated deformation in both asthenosphere and lithosphere.