I was just wondering if there is a way to make the postcode a more prominent aspect of matching addresses.
We run online petitions on our web platform that usually only asks constituents for postcode details rather than full address details (though occasionally we do get full details for actions where they email to their MP) and unfortunately the website database does not overwrite or clear any Address Lines if the postcode entered is different. So when I'm importing the data from these petition actions the address details are incorrect and ImportOmatic does not seem to catch these differences as the Address Lines (and City/County/etc) are a 100% match.
I don't see how the address can be a 100% match if the postcode is completely different but I haven't found a way besides requesting to see all matches during import and filtering through them manually. Our other teams are hesitant to add more form fields as they often hinder completion rates.
Would be grateful for any advice
We have exactly the same issue with data from the Engaging Networks platform. Our best solution so far is to use the IOM API.
If the data is from a form we know to ask only the postcode, we blank the other address address fields in the AfterDictionaries Event. This slightly reduces the overall match rate, but prevents IOM from adding an old address with a new postcode. Because the petitions have to use an email address, the matching does still work pretty well.
You'd need to decide how to use the config setting for IOM to ignore new addresses with no address block. Do you want your existing addresses replaced by a new address with only a postcode?
If you're able to use the API (ie not hosted by blackbaud, and have someone familiar with visual basic) I can give you more details. Otherwise, you might consider pre-processing your data file (in Excel?) to blank the address fields for the petitions in question.
Hope that helps. I'd definitely be interested to hear other solutions if anyone has them.