How to optimize the "creating output data" phase of an on-demand data reload

[Clayton Bridge]

When I reload an on-demand data table, the status starts out as "creating output data" then moves to the "Reading SBDF data" phase. I have two databases that house the same information, and have built two analysis files with the equivalent on-demand data pull based on markings in the analysis to get data from each database. For some reason, when pulling large datasets, one analysis stays on the "creating output data" phase for much longer than the other analysis. They both load the data at about the same rate, but the dashboard that stays on "creating output data" longer takes a much longer time for the data to reload, because of the "creating output data" phase.

Can someone please explain to me exactly what Spotfire is doing during the "creating output data" phase and offer some suggestions on how I might optimize it

Thanks for your help!

Clayton