Has anyone else encounted a large amount of errors in a screening?  We got both Wealth Point & Prospect Point and screened over 100K donors from an initial 1.3M database.

We are still working with the results three years later, but now that we are getting down to doing in-depth research of prospects, I've find a lot of errors in real estate vaulations.  This is BIG because Blackbaud bases the model scoring heavily on real estate.  (We first reported these errors in 2006 and BB told us they couldn't do anything about them).  The issue has been re-opened because we want to do another screening, but Blackbaud will not be a candidate unless they can explain why the errors occurred and what they would do to prevent them in the future.

All I get is noise about the errors being passed along by another vendor to BB, but I say if they want to instill confidence, they would correct them.  Originally they tried to blame the errors on individual county assessors, but then when I pointed out (with the help of my IT background) that different assessors all over the country couldn't have made the same mistake at the same time, then they blamed Lexis/Nexis.  It doesn't stop, and I don't care.  BB sold me the data screening, not L/N, they should make it right.

Has anyone else spotted the same sort of errors, or (horrors!) othe sorts of data errors?