Ranking Factors in Google Maps – Cracking the Code SMX Local

Here is the presentation I presented at SMX Local Mobile:

If you have questions or suggestions to this model of ranking factors let me know. I hopefully will be writing up more thorough report on our research shortly.

Mike Blumenthal

Please consider leaving a comment as your input will help me (& everyone else) better understand and learn about local.
Ranking Factors in Google Maps - Cracking the Code SMX Local by

50 thoughts on “Ranking Factors in Google Maps – Cracking the Code SMX Local”

  1. All I can say is; excellent research Mike! And luckily the collaborators were kind enough to provide you with the necessary datasets to do some thorough investigation! Hooray for the community.

  2. Yay!
    I’m so glad you posted this, Mike. I’ve been waiting.

    This was such fun to work on. I want more.

    I think new areas we could potentially explore would include UGC, trust quality of citation sources (web pages), age of listings (though I’m not sure how we would discover that), etc.

    It was so worth it to do this work. Thanks for putting this up for all to see.
    Miriam

  3. As I read this and my brain started working on simple conclusions as brains do in the face of such things. What I kept thinking is that the best long term strategy for a small business in Maps would be to focus on traditional SEO.

    I’m thinking as time goes by one would expect categories to move to the right in terms of competitiveness. So, one might as well focus to the right of where the category is now for long term results.

    Thoughts on this?

  4. Hi Tim,
    I see what you mean. In fact, I’d say that with what we’re seeing with citations…it may not even really be traditional SEO that’s the strongest signal, but rather, a form of Search Marketing engaged in specifically to get the right citations from the right places (rather like link acquisition, but with a wider latitude of acceptable acquisitions owing to the fact that they can be mentions rather than strict links).

    What do you think about that?
    Miriam

  5. Thanks for publishing that, Mike. I agree with Miriam. We should do some more data sets. I’d continue to use the same format for a bit, just to verify the data.

    While we had over 400 different business records we only looked at 4 data sets. Another set or two might skew the information in each category; (few signals -> many data signals.). We might want to look at data wherein the information is spread around over a larger geography including a city with a lot of square mileage and/or a small state.

    The more data -> the better the research.

    Wonderful presentation.

    Dave

  6. Curses! Mike, that pingback is a scraper site. They scraped by Search Engine Guide article. Just took the whole thing. Grrrr!
    Miriam

  7. Thanks, Mike. There are a couple of scraper blogs that seem to just hog any bit of content that I write on SEG. Annoying, but all too common. I appreciate you removing that pingback.
    Miriam

  8. @tim
    Yes that is a good way to look at it.

    There are some differences that correlate well like business title/category and title tag etc but are unique to local…In Sept 2006 I wrote Does Business Name = Title Tag?.

    Then there are the wider latitude things like what does a local link look like. It needs to include pure geo citations as well as solid link phrase link backs type links. If I were to define the ideal web reference it would be a solid link phrase + a complete geo reference all in one. But I would take citations and linkbacks just the same.

    Then there is the straight geo web stuff that isn’t really clear yet. Do KML, KML widgets and GeoRSS files add rank? Or do they just confirm location and legitimacy?

    While I agree with you that the geo web will tend towards more complete web based info and lots of main stream searches will be defined by that completeness on the right side of the graph…there is still the long tail local searches where more purely local factors will come into play (like on the left side). The strategy for approaching those will be unique to local at least for now.

    Mike

  9. Fascinating research here Mike et al. Would it be possible to get details on the methodology, so as to better understand what the numbers mean (e.g. 80% importance of the centroid, 16% variability… that slide in particular didn’t mean anything for a non-attendee :( ).

  10. We measured 25 data points (some with a scraping program and some by hand) on each of 100 businesses in 4 search categories.

    We then ran a stepwise multiple regression on the results to attempt to create a model that would predict ranking by assessing the correlation to ranking of each factor. For each of the four sets we established which variables had the strongest correlation and which in combination created the best predictors of ranking… thus while we can plot correlation we can’t really measure causality. So while I wouldn’t put too much store in the absolute numbers there is a fair bit of confidence in their relative relations.

    The issue of the centroid was also addressed by Carter Maslan in the interveiw with Erek Enge so we are confident that there is causality in this case.

    I may, if I have time, write up more complete coverage of the study.
    Mike

  11. Take it from someone who ranks well in Google Local for some highly competitive terms like search engine optimization portland oregon and search engine marekting portland oregon. It’s all about combining a good listing and good onsite seo, if you can do those two things you’ll get one of these coveted positions. The caveat, Google has changed its local algo before and there’s nothing stopping them from doing it again.

  12. Really great research Mike, I think that the key factors that I have seen over the years and more recently are:

    -organic position
    -distance from the centroid
    -citations
    -pages mentioning your location
    -telephone number
    -physical address

    It probably also helps to be linked to local sites…

    I also read a thread sometime ago and someone was asking about influence being given to the other search sevices that Google has relationships with, i.e. CitySearch, Restaurant Row, SuperPages etc… Was wondering if this may be influential?

  13. Hello Mike,
    I enjoyed looking at the presentation you included here, but it would be more helpful if you could include the audio of your presentation so that the concepts could be understood better. I didn’t notice any way to listen to the presentation you gave. Is that available?
    Thanks,
    Richard

  14. You have presented an awesome research as a simple beautiful slides.
    Goals of the research like Quantify the critical factors in ranking, Understand what we know for sure, Understand what we think we know and what we need to learn, Establish protocol and questions for additional research, So that we can develop best practice marketing that is honest AND cost effective based on facts were good. I’d seen the link to this post in many other blogs and thereby it sure your research was a very successful one. Congrats Mike, you have done a great thing :)

  15. Wow, very interesting read here. I stumbled upon this while surfing over at David Mihm’s site and have to say I am extremely impressed with the level of detail here. Nice job.

  16. @Nick

    I think the model is essentially sound and still the foundation of ranking that being said I do believe that google has added more ranking factors and prioritized others that are not currently showing on the chart…

  17. I’d like to see someone compile a List of the Top 10 IYP and the Top Data Directories where a new business listing should list take the time to detail their company profile and the pitfalls for not doing so.

    I keep reading that Citations are the new Link for local search and that consistency matters quite a bit when populating a new citation.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Comments links could be nofollow free.