A meeting prior to a data discovery spike tends to use time poorly for everyone involved. In corporate situations (ie: not contract/client based) where you can spike on a data set prior to meeting about the needs, this will save everyone time.
As you first review the data (spike), you can catalog as needed, provided that this isn't already available. If it's available, compare how it's cataloged versus what's actually there (in some cases, data is labeled incorrectly). With this technique, you'll have an understanding of what you're looking at and where the data live. When you meet about the data needs, you'll already have mental links to the data.
This is the point where I'll use pre-built queries/analysis to review the data set.
The result of this is that you'll take much better notes due to the mental links plus you'll ask better questions if you have them. The best part is that you'll save everyone time - there is no shortage of people in the corporate world complaining about the number of meetings they attend!