As my models have become larger and more complex, I've noticed some slowdown in slicer updates, etc. so I've been investigating optimization methods. One of the things I've read is that sorting your data by order of columns with the least variability first improves performance, but only if the sorting is done before you import into PowerPivot. I just want to clarify - Should be able to sort my source data then UPDATE the data in PowerPivot and get the same result?