The fact that many people doesn’t live in a single family house but live in a flat sharing the same building number on a street with people living in other flats in the same building is a common challenge in data quality and data matching.
The same challenge also applies to companies sharing the same building number with other companies and not to say when companies and households are in the same building. So this is a common party master data issue.
Address verification and geocoding is seen as important methods for achieving data quality improvement related to the top data quality pain all over being quality of party master data and aiming at getting a single customer view.
Multi-occupancy is a pain in the (you know) getting there.
My pain
I have had some personal experiences living at multi-occupancy addresses lately.
One and a half years ago I was living a painless life in single family house in a Copenhagen suburb.
Then I moved closer to downtown Copenhagen in a flat as mentioned in post Down the Street.
The tradition in Denmark is to send letters and make deliveries and register master data with a common format of units within a building and having separate mailboxes with flat ID and names for each flat. I have received most of my post since then and got all deliveries I’m aware of.
Then I moved to London in a flat. Here the flats in my building have numbers. But the postman delivers the letters in one batch in the street door, and there are no names on the doorbells in front of the door.
So now I sense I don’t get many letters and today I had to order the same stuff trice from amazon.co.uk, because I haven’t received the first two packages despite of their state of the art online accessible package tracking systems that tells me that delivery was successful.
Master data pains unresolved
Address reference data at building number level and related geocodes are becoming commonly available many places around these days.
But having reference data and real world aligned location and related party master data at the unit level is still a challenge most places. Therefore we are still struggling with using address verification and geocoding for single customer view where a given building number has more than a single occupancy.
Liliendahl on Data Quality 
Hi Henrik
From what you write it seems that there is no problem with the data structure or granularity as it is capable of defining a unique Delivery Address for a Party, i.e. Henrik Liliendahl Sørensen, Flat 6, 130 Nice St, London. NW72AJ.
What is wrong is the delivery mechanism which takes uniquely addressed items and removes the Party Name and the Flat Number by bundling them at Street Address level, thus removing two levels of granularity. No additional data would overcome this flaw in the delivery mechanism, compounded by the fact that someone may well be stealing the mail.
As a data quality practitioner you should perhaps be consoled that the level of granularity not only allows the thief to know precisely who he is stealing from but also what you like reading, from whom you buy it, the frequency of purchase of specific titles and the cost of each item.
Regards
John
Thanks for commenting John.
My current delivery pain is most probably due to theft made possible by that fact that my mail and package delivery point is organised as if we are one household in the building which probably was so in the old times with the premise being an upstairs/downstairs residence.
I actually had great difficulty registering the address on amazon.co.uk as the address verification clearly wasn’t aware of the address being a multi-occupancy and which units actually being there.
Somehow such circumstances seem to occur together.