• http://silvery.com Chris Silver Smith

    I think the data must be slanted by Chitika’s relatively limited set of sites to derive data from. I don’t believe these findings can be representative. In smaller towns, there are fewer providers, so there’s actually less reason to perform local searches to find more service providers. I think there is actually less online local searching going on in small towns, which is partly why print directories are projecting those markets to remain stronger areas for print yellow pages products to remain viable. Small, rural towns in the South also have slower average internet access speeds, IMHO, which also discourages online searching.

  • http://bulwarkpestcontrol.com Thos003

    Hmm… Where are they getting this from?

    The possible flaws in this logic. If they are considering searches with local keywords then I could see this conclusion. Most large cities have better developed local search and businesses involved in SEO. Therefore while in Phoenix one could simply search for “pest control” and get the results that they want without having to refine it to “Phoenix pest control”. If they miss the fact that both searches have local intent then the survey will be flawed. I believe many people and local qualifiers to search after being disappointed with the first search. This also could be included out of habit for smaller towns simply because of past experience.

    Do their numbers add up with what Fredrick Vallaeys reported? – http://twitter.com/#!/Thos003/status/63419377034334208

    Could this be based solely on desktop searches? Are more city dwellers carrying smart phones?

  • http://pgabor.com GaborPor

    It’s unfortunate that the Chitika article didn’t cover how they defined “local” search query. Considering that Google is aware of the (approximate) location of the searcher (based on his/her IP) searching for “car service” should also be considered local, just like “car service in Boston.”

  • Matt McGee

    Gabor – did you read the complete article above? I list 15 example queries that were counted as local and it includes several phrases like the one you’ve mentioned here.

  • http://www.vikchhabra.com Vik Chhabra

    Chitika also did a study that showed that most local searches are performed on mobile devices –> http://insights.chitika.com/2010/mobile-users-45-more-interested-in-local-than-non-mobile/. I would like to see what percentage of those people in the smaller towns listed have smartphones vs. the larger cities listed.

  • http://bulwarkpestcontrol.com Thos003

    …sorry… pest control guy typo…

    “I believe many people add local qualifiers to a search after being disappointed with the first search.”

    BTW… if they count both “pest control” and then the follow up search “pest control Chattanooga” as 2 separate searches then their numbers would also reflect a higher searches for those areas where local search is not as developed.

  • http://www.gabedonnini.com Gabriel Donnini

    Hi everyone, Gabe from Chitika here.

    The data we reported was in the form of ratios, so, many of you are correct, that there are fewer people actually “searching local” in the more rural areas compared to cities.

    The purpose of this study is not to find out where the largest raw number of people are searching local, rather, it is to find out where the greatest percentage of traffic is composed of local queries. The keywords in our algorithm are not location specific in any way.

    Although we give preference to queries with location in them, our algorithm primarily consists of local categories which you might find in a phone-book.

    If you have any other questions, check out our website insights.chitika.com or shoot us a message on Facebook.

    Thanks! – Gabe

  • http://www.betterfindability.com Clark Mackey

    Thanks Matt and Chitika for the thought provoking post.

    Right off the bat, only 1 of the cities mentioned is small and none are rural (rural was not mentioned in the article but it mentioned in comments). Several are Southern. I think the explanation is partly related to the very real-world movement afoot to buy local. Lexington, Greenville, Knoxville, and Chattanooga fit the bill particularly well – relatively forward thinking municipalities with a well-off population committed to improving the local economy. Add to that a dash of Southern-ness and there you have it. More local search. And that is cool.

    To be fair it is also easier to isolate IP addresses and serve the right local results in towns of this size than say, L.A. In a larger market knowing your IP address doesn’t tell you enough about whether to serve a local result in the way that it does in Greenville, SC. I see more errors served – failures to deliver good local results – in these confusing larger markets. Examples of confusing IP addresses that may or may not give you the right take on local: West Hills and L.A.; Chula Vista and San Diego; Evanston and Chicago; and Marietta and Atlanta. I’m sure there are lots of other examples of areas where IP address can fool the search engine in either direction, either into showing an unnecessarily limited local result or into not showing one at all. In these markets IP filtering is only part of the answer (the other parts are user intent and behavioral patterns). I think getting local search right is more difficult in the larger cities and that may be resulting in the appearance of a smaller percentage of local search.

    The outlier in this group is Pensacola FL (56k population, beach town, military town). The others are all progressive mid sized cities 300k + in population.

  • http://www.freedomvoice.com Nick Gowdy

    My first thoughts for explanations are:

    1) Population density. Like you touched on at the end of the article, people in larger (denser) cities presumably don’t have to travel as far to find things. You step out of your building and within a five block radius you have just about every kind of business you could imagine.

    2) Real World Advertising. Walk through the heart of some of these bigger cities, and you’ll see ads on the streets and on buildings, on cars and buses, people handing out fliers, big neon signs, etc. You know where things are because you’ve been subjected to an all-out advertising blitz on your corneas for as long as you’ve lived there.

    3) Yelp & Social Review. Maybe these larger cities are the “high-tech hubs” we think they are, but that means they’ve moved their search for services like those listed (restaurants, car repair, laundromats, etc.) away from major search engines. Now they’re tweeting, “Anyone know a good plumber?” or they’re hopping on Yelp to find the best breakfast burrito in the Mission.

    4) Maybe large cities just do a lot more non-local searching? It’s a shame these are percentages of searches, not number of searches vs. population size. I’d be inclined to believe that, even if 1 – 3 are wrong, perhaps the same number of local searches are being performed per person in these cities, but that these cities do a ton of non-local searching that skews the ratio.


  • Dave Oremland

    Real interesting results. It would be interesting to fill in the data with percentages of searches for other broad categories beyond local search. How does it all break down? Another way to look at the data would be to compare total search phrases to a \population norm\ roughly equal to the population in that metro region. Then establish a number that roughly equals the # of searches per person in that region.

    Just suppose people in the regions with relatively fewer local searches by percentage actually made more searches per day, if one used the above analytical method. Further research might suggest that the rough number of local searches per person might be more equal than that first presented.

    Or it might not. :D Its provocative information. Chitika should further break it down to provide more insights.


  • http://screenwerk.com Greg Sterling

    I don’t believe these data are accurate. Local search volumes should be relatively consistent across markets. Overall volumes should track population generally speaking.

  • http://www.ericward.com Eric Ward

    The appearance of Knoxville as #3 is skewed…by me. I do a few hundred local searches a day for Link Building and SEO Research, and I live there. :)

  • Stupidscript

    The question GaborPor posted is valid, Matt. There needs to be a definition of what Chitika means by “local”. That definition is not at all clear in either your article or in Chitika’s.

    ALL search is “local” due to the use of geolocation metrics and other methods (e.g. “personalization”).

    So what makes these searches particularly “local”? The only defining “local” element (absent those imposed by the engines) in the examples given is the use of a location in the query.

    Searches for “car service” are NOT inherently “local”. The results displayed will automatically be localized by the engines due to the use of the metric/method noted and others.

    Perhaps Chitika means that the searchers RESPONDED to local results … i.e. by clicking one of the Places links, or by moving to the Maps links, or by triggering one of the “Local” PPC ads?

    Can’t tell without knowing how Chitika defines “local”.