I've been working on an old events database that lists delegates from years back. It tracks who's been to previous events. The database was built back in the faxback days of manual entry and I'm trying to update the workflow.

 

My idea is to get a data dump from the events registration web backend and try to import/match as many records as I can. Some things I may be able to match are mobile/cell phone number, firstname-lastname joined, or browse address. The idea being I have an intermediate import and cleanup file before introducing the data into the main database. Are there any other ways of achieving this? It would require manual identification before each record gets the to-import tick.

 

Also, how do you keep track of what has been imported from the web backend, to avoid importing multiple times? They don't want to delete the entries from the web backend.

Have a flag in the intermediate database to indicate a match has been made and pushed into your final database. These records can be ignored on any subsequent passes.

 

And another flag to indicate a possible new match, for the next push into the final database (or import from intermediate to final database).

 

This should allow you to try different match keys, and then ignore records that have been matched previously.

 

I'd probably also have the final database primary key able to be entered into the intermediate database to procide a match / update field into the final system.