Does anyone know how Washington Post has managed to source Google Trends data at county level for this map?

enter image description here

Here's a state-level query for Kansas. The geographical response is by metro.

enter image description here

In the (unofficial) API response this data is labeled interest_by_dma (designated market area), which is more granular than state level, but less than county as per Washpost's map.

I've heard that Google does not sell Trends data. Assuming this is right and Washpost does not have some private arrangement, does anyone know how they might have sourced this data?

  • 1
    Very similar picture has been "created by the Google News lab using CartoDB [and] published on August 1, 2017": 1, 2. – Stanislav Kralin Aug 5 '17 at 13:57
  • Thanks @StanislavKralin. I'd hazard a guess the carto map is exactly the same source data with 7 classes instead of 10. – geotheory Aug 7 '17 at 17:53

While I don't work for The Post or Google, and thus cannot speak authoritatively, I was at a Google event for top news organizations last year in NYC, and can attest to the fact that -- depending on the publication -- Google often provides custom Trends data as an enterprise service to publishers. That would be my guess here.

  • I think you're right - my understanding of Google's policy was wrong. Clearly some arrangement with (at least certain) news organisations, but not a service open to all. – geotheory Aug 7 '17 at 17:46
  • Yes, Google has a tense but symbiotic relationship with news orgs generally, as it's largely their original content which is the chief driver of search. So they have special programs to keep publishers happy and to encourage them to create AMP versions of their content. – Carl V. Lewis Aug 7 '17 at 17:52

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.