Recently, Clearly Contacts has a promotion where they are giving away 500 pairs of glasses for free in several cities across Canada. Being a freebie, this deal is quite popular with RFDers and some members have claimed to get up to 5 free glasses in similar CC promotions in the past!
Of course, CC has a policy of only giving one free pair of glasses to a household, and I expect they have some sort of automated way to determine duplicates (because some people get caught). But, I was thinking about it and it is not straightforward to determine which orders are dups while limiting false positives.
The obvious is to use street address, but that quickly falls apart when you consider apartment residents with the same street address. You can’t allow variability based on apartment number, because then a (single) household can abuse that field. You have the same problem with postal code.
You can try to match on personal information such as email address, credit card number or name; everyone has multiple email addresses, most have several credit cards, and you can modify your name and the package should still get delivered to you.
The last piece of information is your prescription. I don’t know how statistically similar prescriptions are, but I think that needs to be the key. If you can find multiple orders with very similar prescriptions, and some other information (such as name or email or street address) are statistically similar, then those are probably orders from the same person. At least that would be how I would implement it.