Different locations recorded in AdWords reports and third party tool
I am running a campaign targeted to specific regions in Germany tracking results both in Google AdWords reports and Supermetrics Add-on for Chrome. While the overall figures for clicks by country are the same in any of these reports for a given period of time there is substantial differences in user location reporting. Some locations (cities) are recorded in Google AdWords only, others show up in Supermetrics reporting only.
I asked Supermetrics support and they said there is nothing to be done about it because that's how they get the data from AdWords API.
Is Google using different database sources with their AdWords service than with AdWords API? How is the user location dtermined anyway? Just by IP address or by other means?
Thanks in advance and kind regards,
Re: Different locations recorded in AdWords reports and third party to[ Edited ]
December 2014 - last edited December 2014
Welcome to AdWords Community.
As far as I know, data sources are the same for AdWords and its API. There shouldn't be any difference in the reports.
But yes, how that third party tool is coded to get the data is important aspect here. If you ask me, I would believe more in the AdWords report than in any third party tool for this reason only.
To your other question, you may refer to the following URL to know how location is determined -
The related documentation for API is here - https://developers.google.com/adwords/api/docs/guides/location-targeting
Join me on Google+
I agree with Ratan on this. I would go with the AdWords Geo reports any day over any third party reports purely because it is Google's algorithm that allows AdWords to pin point a users location.
As for the third part reports using AdWords API, it could be that the location metric/column that they are pulling from the AdWords report to prepare their own version is the correct one.
Basically, if you pull a geo report in AdWords, you will see levels of geo details for every click, impression, etc. starting from Country, state, city to the most specific user location. Now since there are so many levels of geo details available, it could very well be the case that the Supermetrics algorithm pulls a different geo detail amongst the ones available using API's as compared to the AdWords report that you pull and the geo detail that you choose to report.
Hope this Helps!
Shashank Singh, AdWords Rising Star |
Did you find any helpful responses or answers to your query? If yes, please hit "Like" or "Accept as Solution"
I agree with @Ratan-Jha on this.
3rd party tools use the Adwords API. So the source is the same.
The difference could come in the format of the report. Note that Adwords breaks down Geo reports into two segments. People physically in the region and people who search about the region. (Search location by intent)
Now think of a user in Munich submitting a query which includes Munich as a location. How this is counted in the report? As a user physically in Munich or as a user who runs a "search intent query"?
Did you find any helpful responses or answers to your query? If yes, please mark it as the ‘Best Answer’
Re: Different locations recorded in AdWords reports and third party to
thank you very much for your helpful comments and links. With conflicting info on locations in the reports I guess I have to stick with Google as long as I don't know how data is retrieved. Nevertheless, I regard Supermetrics a great tool to supplement AdWords because of the ease it gives to visual reporting.
Moshe, I thought about the concept of user location vs location of interest, too. I checkmarked user location and I think the mismatch has a different reason. I would expect way more clicks from other regions in the country if Supermetrics would gather the figures from the location of interest option. On city level the mismatch may be more than 2% of clicks per month while on region level (German provinces) it may be 0.4 - 0.85 percent only. I am saying that the clicks are coming from broadly the same or adjacent region although the towns and cities recorded in either report may be 150 kilometres apart. So to me it still looks like a difference in IP addresses and/or celltower ID data.