Studying mantle convection requires knowledge of how plates moved over and subduct into the mantle. Therefore, relative plate tectonic reconstructions are placed in a mantle reference frame. These use the geological expressions of plate-mantle interactions and correlate these to mantle structure or minimize plate motions that would cause friction with the mantle under the null hypothesis that active horizontal flow in the mantle is negligible. However, mantle reference frames based on different plate-mantle interactions are different. This may be due to model uncertainty, but may also hold meaningful geophysical signals. To explore this, we first computed a reference frame, in 10 Ma steps, that adheres to a ’tectonic rule’ that minimizes absolute total continent motion. We estimate the uncertainty by assigning a ±5 Ma time window to the 10 Ma intervals and find that the continent frame, or alternative frames based on an alternative ’rules’, may provide meaningful results for the last 350 Ma, but are unresolved before that time. With this base frame, we predict hotspot tracks, kimberlite and large igneous provinces eruption sites, net lithosphere rotation, trench kinematics, and true polar wander, which are all mostly within plausible ranges. We introduce this coupled frame as a Solid Earth Integrated Reference Frame (SEIRF) that may be used (1) to aid interpretation of anomalous geodynamic behavior; (2) to obtain novel constraints on mantle convection - the SEIRF allows studying ’mantle kinematics’ in a plate tectonic reference frame and (3) may serve to train 3D numerical models of solid Earth dynamics.