Google Trends on the Real Estate Crash

One of the more interesting features of Google Trends is that you can see a normalized table of where searches are coming from. For example, if you check Google Trends for searches for “real estate crash” one result is the following list of the top ten cities from which searches for those keywords have been originating:

Look familiar? It should. Google has back-doored its way into a reasonable cross-section — call it the “residents’ list” — of the most over-priced real estate markets in North America. Put differently, Google becomes one good way of finding out where the real estate natives are most restless.


  1. great find..great way to analyze…
    amazing what all google can do…

  2. Do they adjust the data for size? I.e. do they adjust for searches per thousand or do they only go by gross number of searches?

  3. Ben — That’s the intriguing thing. All the search data is normalized by Google first:

    oogle Trends first looks at a sample of all Google searches to determine the cities from which we received the most searches for your first term. Then, for those top cities, Google Trends calculates the ratio of searches for your term coming from each city divided by total Google searches coming from the same city. The city ranking you see on the page and the bar charts alongside each city name both represent this ratio.

  4. Very neat and looks like a powerful tool.
    I also quite like the news items that are provided by the trends analysis tool. An example is searching for “calgary flames” will show some news articles on the side. One problem with these news links are that many of them are invalid links as the articles had been removed.

  5. How do we know this just isn’t the Geo-location of the Kedrosky readership?
    We haven’t seen one of those nifty correlation graphs lately. Look to the log files!
    BTW, when are we gonna see the defintion of “crash” and “bubble”. Any quantifiable meaning here?

  6. John —
    Geolocation of the Kedrosky readership? Sadly, I have no influence on how Google constructs its trends graphs. And in scanning my location data, most of my readership is in Kazakhistan.
    Wrt definitions, are you suggesting that U.S. real estate has not recently experienced dramatic price appreciation, something that might loosely be described as a bubble?
    While Investopedia and others offer definitions, and my view of bubble and crash is mostly significant moves (like +/- 20% in shorter than historically normal periods), you won’t catch me pretending that these words have Platonic solidity.

  7. Ah, but you do have *some* influence on Google trends – you just searched for “real estate crash” and probably caused some others to do the same.
    And the coincidence of San Diego being #1 and also your usual geo-location? Hmmmm… (Raised eyebrow)… Might you be over-influenced by local conditions?
    Re: definitions –
    Overall, yes, I believe you over-use “bubble”. But that’s part of your allure too :)
    I think bubbles are defined after the fact, because it takes at least a 50% decline to pop the bubble. Have we seen anything like that?
    In fact, where are we seeing the -20% in a “shorter than historically normal” period? If anything, that seems like a correction to me.
    So I generally think “market” when you write “bubble”. All in good fun, though.

  8. Notable, by its absence, is silicon valley. Search “Real Estate bubble” and valley towns show up in spades. Shows geographical nomenclature diversity. In silicon valley crashes are for computers!

  9. Do you know where I can get more recent stats on this kind of info?

  10. See here.