A very interesting article, I'm looking forward to the follow ups as we're heading in to a fairly big matching job (merging 1/4 million education records to 100k social care records for our county).

A couple of questions (which you might be covering in the future). Does fuzzy matching work in the same way as fuzzy grouping? I had assumed that there was some blocking going on in the background on any exact matches defined but the tests in the article imlies not.

Also is there a way of getting the blocking to work in SSIS with a much larger number of blocks e.g. one for every date of birth or post code? We could probably run the fuzzy matching in a loop but that would loose the advantage of the blocks running in parallel.

Thanks

Barney