Finding unmatched zip codes where address matches

Question

I am trying to use Access to go through a large database of mailing addresses, and I'm looking to generate a report that looks at all of the mailing addresses that are identical, and then looks for zip codes that do not match. Ideally, I would like to use this same method to check Names vs Mailing Addresses, Mailing Addresses vs Names, etc.

The goal is to clean up the database and drastically reduce postage.

I had thought this was going to be fairly simple, but I haven't come up with a way to do it yet, and I haven't found anything online that tells me anyone else has done this.

My current attempt looks like this: but it is returning zero results.

SELECT [Permissive Export_OLD 1].ACCOUNTNO, [Permissive Export_OLD 1].KEYNO, [Permissive Export_OLD 1].NAME1, [Permissive Export_OLD 1].NAME2, [Permissive Export_OLD 1].ADDRESS1, [Permissive Export_OLD 1].ADDRESS2, [Permissive Export_OLD 1].CITY, [Permissive Export_OLD 1].STATE, [Permissive Export_OLD 1].ZIPCODE
FROM [Permissive Export_OLD 1]
WHERE ((([Permissive Export_OLD 1].ADDRESS2)="EQUAL") AND (Not ([Permissive Export_OLD 1].ZIPCODE)="EQUAL"))
GROUP BY [Permissive Export_OLD 1].ACCOUNTNO, [Permissive Export_OLD 1].KEYNO, [Permissive Export_OLD 1].NAME1, [Permissive Export_OLD 1].NAME2, [Permissive Export_OLD 1].ADDRESS1, [Permissive Export_OLD 1].ADDRESS2, [Permissive Export_OLD 1].CITY, [Permissive Export_OLD 1].STATE, [Permissive Export_OLD 1].ZIPCODE;

Answer 1

I recommend the following query:

SELECT [Permissive Export_OLD 1].ACCOUNTNO, [Permissive Export_OLD 1].KEYNO, [Permissive Export_OLD 1].NAME1, [Permissive Export_OLD 1].NAME2, [Permissive Export_OLD 1].ADDRESS1, [Permissive Export_OLD 1].ADDRESS2, [Permissive Export_OLD 1].CITY, [Permissive Export_OLD 1].STATE, [Permissive Export_OLD 1].ZIPCODE
FROM [Permissive Export_OLD 1], [Permissive Export_OLD 1] AS [Permissive Export_OLD 1_1]
WHERE ((([Permissive Export_OLD 1].ADDRESS2)=[Permissive Export_OLD 1_1].[ADDRESS2]) AND (([Permissive Export_OLD 1].ZIPCODE)<>[Permissive Export_OLD 1_1].[ZIPCODE]))
GROUP BY [Permissive Export_OLD 1].ACCOUNTNO, [Permissive Export_OLD 1].KEYNO, [Permissive Export_OLD 1].NAME1, [Permissive Export_OLD 1].NAME2, [Permissive Export_OLD 1].ADDRESS1, [Permissive Export_OLD 1].ADDRESS2, [Permissive Export_OLD 1].CITY, [Permissive Export_OLD 1].STATE, [Permissive Export_OLD 1].ZIPCODE;

This query compares each row in the table against every other row in the table to find ADDRESS2=ADDRESS2 but ZIPCODE<>ZIPCODE. The query will return all records which fit that criteria as separate rows.

Answer 2

Have you considered an address correction service like what SmartyStreets offers? (I work at SmartyStreets.)

Finding duplicate addresses is exceptionally tricky without putting them through a validation and standardization service. Such services can handle mis-spellings, addresses in different formats, etc. Your query expects the same addresses to be exactly equal and won't match "123 Main" with "123 Main St" even though they're the same.

And since these services tend to be CASS-Certified, assuming you have any US addresses, you'll get a CASS Summary Report which you present at the post office for lower mailing rates.