Difference between Country and geo_country column in Data Feed

Description

  • Below is Sample Raw Data for such event:
browser code_ver country geo_city geo_country
0 JS-2.10.0 32 shibuya jpn
2572806621 JS-2.10.0 0 okinawa jpn
2792242630 JS-2.10.0 304 misawa jpn
0 JS-2.10.0 32 shibuya jpn
2792242630 JS-2.10.0 0 osaka jpn
1045425661 JS-2.10.0 300 ota jpn
2792242630 JS-2.10.0 300 tokyo jpn
2792242630 JS-2.10.0 300 tokyo jpn
3754824564 JS-2.10.0 0 okinawa jpn
2792242630 JS-2.10.0 300 tokyo jpn
2792242630 JS-2.10.0 300 tokyo jpn
2572806621 JS-2.10.0 304 tokyo jpn
3754824564 JS-2.10.0 0 okinawa jpn
1951947700 JS-2.10.0 45 shinagawa jpn
3754824564 JS-2.10.0 304 shibuya jpn
0 JS-2.10.0 32 shibuya jpn
0 JS-2.10.0 32 shibuya jpn
2792242630 JS-2.10.0 300 tokyo jpn

Resolution

The country column is actually the top level domain.  In the “beginnings” of the internet, the top level domain and the country were basically the same thing, which is probably why the column is called country.  There was no IP to geo info mappings at that time. If the country column in the data were to be named now, it would be called top level domain (or something like that, but not country).

With IP to geo info mappings that are used now, the IP of the hit is used to determine the geo_country, geo_region, geo_city, etc. as well as the domain.  The top level domain (the country column) is determined from the domain.  The domain, however, has become less “significant” in the IP geo info mappings.  In fact, the domain is no longer required to be provided when IPs are purchased/allocated/etc.  It is up to the IP block owner to determine if they provide domain info, and if so how detailed/accurate information they provide. With many international companies, etc. now, even when domain information is provided, they will often provide the same domain for all IP blocks used throughout the world (which means the top level domain won’t match the geo_country, or the country where the IP is actually used).

So, in Adobe Analytics, domains such as au-net.ne.jp, ocn.ne.jp, spmode.ne.jp, transix.jp, etc. (those where the domain ends in .jp) will show as Japan in the country column.  While domains such as bbtec.net, jptransit.net, ntt.net, etc. (those where the domain ends in .net) will show as 304 (Network) in the country column, those domains such as aetosjapan.com, zscaler.com, etc. (those where the domain ends in.com) will show as 300 (Commercial) in the country column, etc.

As the domain has become less significant in IP mappings, and as domains with .net, .com, etc. have become much more common outside of the US, using the top level domain (country column) is no longer as valuable for determining actual country information for the origins of the hits for a visitor. The geo_country column is much more accurate for determining the country of origin of the hits for a visitor.

One additional thing to also remember, although this only applied to one of the hits with geo_country = “jpn” for that day, but can be significant especially if traffic is from mobile devices, and other situations, that may “change” IPs mid visit.  In Adobe Analytics to provide consistent reporting throughout a visit, for certain technology, etc. fields (country, domain, geo_country, geo_city, browser, os, etc.) the value from the first hit of a visit is used for every hit in a visit.  This can sometimes cause issues in debugging when looking only at the IP for a hit and wondering why in Analytics a different domain, geo_city, etc. is used than the IP would resolve to.  The values from the first hit of a visit are used for all hits in the visit.

On this page