M
M Morgan
I gave some data to a call center to do a record update.
These people had NO clue about how to handle customer data..
so, over the course of 10 days, they made multiple calls to my list, and I
now have 10 tables (one for each day), containing a lot of duplicate
records.
The only field that is different on these records is called results. I
obviously care more about some results that others. valid is a much better
response than busy, or no answer or 12 other possibilities.
something like
5551231234 08/12/05 busy
5551231234 08/13/05 noanw
5551231234 08/20/05 valid
5559999999 08/13/05 disconnected
5557895555 08/12/05 noans
5557895555 08/13/05 wrongnum
would all boil down to
5551231234 08/20/05 valid
5559999999 08/13/05 disconnected
5557895555 08/13/05 wrongnum
So, trying to consolidate this all back to my master list, I've reach a
stumbling block.
I've piped everything back into one table, updated the results to as few as
possible (15 distinct results)
My first thought is to create a new table for each of the results, then
update the master by querying each in turn, starting with the lowest
priority result. ( and no, date can't be considered valid, unfortunately)
I have no idea how to loop through a make table query so each new table is
named by it's result code.
I could of course do it by hand, but then I wouldn't be learning anything,
and I'd still be doing it by hand the next time.
I might also be trying to re-invent the wheel here. something as simple as a
group by? I just don't know.
So, TIA for any light that might shine upon this post
These people had NO clue about how to handle customer data..
so, over the course of 10 days, they made multiple calls to my list, and I
now have 10 tables (one for each day), containing a lot of duplicate
records.
The only field that is different on these records is called results. I
obviously care more about some results that others. valid is a much better
response than busy, or no answer or 12 other possibilities.
something like
5551231234 08/12/05 busy
5551231234 08/13/05 noanw
5551231234 08/20/05 valid
5559999999 08/13/05 disconnected
5557895555 08/12/05 noans
5557895555 08/13/05 wrongnum
would all boil down to
5551231234 08/20/05 valid
5559999999 08/13/05 disconnected
5557895555 08/13/05 wrongnum
So, trying to consolidate this all back to my master list, I've reach a
stumbling block.
I've piped everything back into one table, updated the results to as few as
possible (15 distinct results)
My first thought is to create a new table for each of the results, then
update the master by querying each in turn, starting with the lowest
priority result. ( and no, date can't be considered valid, unfortunately)
I have no idea how to loop through a make table query so each new table is
named by it's result code.
I could of course do it by hand, but then I wouldn't be learning anything,
and I'd still be doing it by hand the next time.
I might also be trying to re-invent the wheel here. something as simple as a
group by? I just don't know.
So, TIA for any light that might shine upon this post
