A computer-implemented method includes: determining a reading performance of a head at a commanded lateral reading location; adjusting a location of the head relative to a medium by moving the head in a first lateral direction to an adjusted lateral reading location; determining a reading performance of the head at the adjusted lateral reading location, and determining whether the reading performance at the adjusted lateral reading location is better than the reading performance at the commanded lateral reading location. In response to determining the reading performance is better at the adjusted reading location, the method includes iteratively repeating the adjustment of the location of the head until the reading performance of the head at a current iteration is worse than the reading performance of the head at an immediately previous iteration. Corresponding systems and computer program products are also disclosed.