Ay Caramba! Case of the Missing Tasty Mexican

Call me a local nut, but sometimes I just sit around looking at listings of local businesses…for fun. Tonight I came across a rather poignant example of why it’s so necessary to claim your Google Local Business Listing.

There’s a Mexican restaurant in the San Francisco Bay Area called Maya Palenque. I’ve always liked the name. They’ve been in business for years and years and I imagine they have quite a few regulars, but I doubt they’d be happy to learn that any new potential patrons are being sent to Texas by Google when looking to dine with them. Here’s their listing in Google, with the website of Texas restaurant called the Palenque Grill being incorrectly billed as Maya Palenque’s home base on the web:

Maya Palenque incorrect Google listing

The correct URL for Maya Palenque is TastyMexican.com and I’ll shoot them a little link juice in case it may help Google hit upon the wide difference between Texas and California. Sadly, the restaurant’s website isn’t in very good shape and is lost on Page 2 of the Universal SERPs for its own name (it’s not even ranking at the top for Tasty Mexican…I wouldn’t have thought that would be terribly competitive…or terribly useful!)

Looking at the citations for the restaurant, we see a big culprit right at the top of the list:

Maya Palenque incorrect Google listing

For some reason, the completely unrelated grill in Texas is being pulled as a citation for the California restaurant. There is no connection I can see. It’s not a chain. It’s not even the same name. This isn’t a case of Bob’s Doughnuts in Phoenix being confused with Bob’s Doughnuts in Chicago (oh, THAT Bob’s Doughnuts!). I haven’t looked very deeply into how Google’s signals got mixed up on this but it’s just one more sterling example of why not claiming your listing is dangerous and, as my Mexican great-great grandmother would have said, “estupido!”

But Abuelita didn’t have the compassion I have for restaurant owners like the head man at Maya Palenque who is doubtless without a clue that his guests are being told to go eat in Texas by Google. Really, what I’d like is to help the guy out. In just minutes, he could be logging in, fixing that web address and getting his online rep back on track. And then I’d like to get my little hands on his website!

But speaking of that, how much attention is Google really paying to websites when it comes to Local? My gut feeling is that your website is very important, but then I see Maya Palenque ranking near the top of Maps for a broad search for ‘restaurants novato ca’ and it’s clear that their 30+ citations, 40+ reviews and other factors are outweighing the fact, in the mind of Google, that their restaurant is apparently located way down south in Laredo, TX. Pretty strange.

In conclusion, I hope someone who eats or works at Maya Palenque will stumble across this post at some point and take strong action to clear up this situation that is doubtless bewildering hungry diners across the Bay Area and beyond. Buena Suerte!

———-
UPDATE: As of February 8, 2009, Maya Palenque’s Maps’ lisiting has been fixed to show the correct URL: tastymexican.com. I don’t know if this post had anything to do with it, but I’m glad to see this change!

21 Responses to “Ay Caramba! Case of the Missing Tasty Mexican”

  1. on 19 Jan 2009 at 9:27 am emad

    A lot of the data in Google’s Maps product is from sites providing them data feeds (as well as base data providers). Chances are that one of the sources of this data got it wrong. You can hunt down other sies referencing Maya Pelenque and using the wrong link and send in a correction.

    FYI, I am a cofounder of YellowBot and we also supply Google with a data feed. I have updated this listing:
    http://www.yellowbot.com/maya-palenque-restaurant-novato-ca.html

    The URL will be exported to Google in our next data fed to them but it is entirely possible the incorrect record can somehow take some sort of precendece over our record. I hope this helps! :-)

  2. on 19 Jan 2009 at 9:31 am David Mihm

    Very, very interesting post, Miriam. Particularly the lack of importance that the linked-to website seems to play.

    In general, I’ve noticed that Yahoo Local seems to require more specific business name data in order to match 1:1…do you know if they’ve managed to avoid this mistake? And I should also say that in general I like Google doing close-match stuff like this, i I don’t know the exact name of the business. But this is one example where it does more harm than good.

  3. on 19 Jan 2009 at 3:05 pm admin

    Welcome, Emad, and thanks for making the YellowBot correction. Right on!

    You are quite correct that anyone taking on the task of fixing this error would need to search out any and all references which are confusing the 2 restaurants. That’s exactly what would need to be done and is very good advice. I appreciate you stopping by and, especially, the fact that you took such friendly and prompt action on this. Well done!

    Miriam

  4. on 19 Jan 2009 at 3:09 pm admin

    Hey David,
    As I recall, the restaurant didn’t have a Yahoo listing…or at least, they hadn’t claimed it. I’m glad that strikes you as interesting that the website appears to play so little part in this. The Maya Palenque site does have the business name on the homepage (though not in the title tag) and does have an address, but it is a bit lost in the SERPs and is somehow escaping Google’s notice.

    This is just the sort of project that tempts me to phone the business owner, but telemarketing is not my idea of a good time :) Still, I’d like to see the restaurant understand what is going on.

    BTW, David, CONGRATULATIONS on your Semmys nominations! Yay! I’m so excited about that. I love the Semmys.

    Miriam

  5. on 19 Jan 2009 at 3:41 pm emad

    Thanks, Miriam. I’m not new here, though. I’ve been subscribed to the SEO Igloo Blog since you launched. Your blog is in my RSS reader too, David. Keep up the good work! :-)

  6. on 19 Jan 2009 at 3:48 pm admin

    Well, I’m honored to know that! Thank you, Emad. It’s a pleasure to have you here.

    Miriam

  7. on 21 Jan 2009 at 1:35 pm Martijn Beijk

    Hola Miriam!

    Buen blog tienes aquí, tienes el verdadero espíritu!
    Tu has probado envíar un email a el tasty mexican o no?

    hablamos pronto!

  8. on 21 Jan 2009 at 3:33 pm admin

    Hola Martijn!

    Es favor que usted me hace! Gracias.

    Espero que, tal vez, alguien que trabaje en la Tasty Mexican va ver mi blog post. No me gusta estar como una telemarketer.

    Como va todo en Espagne hoy dia, Martijn? Su facilidad con la lengua es mas, mas mejor que mi facilidad. Lo siento. Comprendo la lengua quando leo y escucho, pero escribir es muy dificil!

    Hasta luego!
    Miriam

  9. on 22 Jan 2009 at 12:02 pm Maps Guide Jen

    Hi Miriam,

    This is one of those great examples where the community edit function is this restaurant’s best friend. If you click “edit” > “edit details” you can correct this business’ website yourself without having to dig through feeds or other references to make a correction.

    Jen

  10. on 22 Jan 2009 at 4:00 pm admin

    Good afternoon, Jen!

    It’s always a treat to see you here. Your point is well taken. It would be a kind thing to do to edit this business’ data for them and I may end up doing that when I have some free time, but I think it’s so much more important that the restaurant owner LEARN that this is going on with his listing so that he can become responsible for his online reputation.

    I like to see small business owners empowered with this kind of knowledge. Self sufficiency can be such a good thing. I’m going to wait a bit to see if one of the restaurant’s employees or patrons comes across this article and lets the owner know that he’s got some work to do online. I would, in fact, be happy to walk him through claiming his listing for free, if he should end up writing or phoning me.

    Thank you for stopping by, Jen! BTW, please pass on my compliments to everyone at Google for the great Martin Luther King Jr. logo this past week. I really appreciated that.

    Miriam

  11. on 22 Jan 2009 at 6:12 pm Mike Blumenthal

    The assignment of the incorrect URL to listings in Maps is a bug that goes way back.

    I would be curious to hear from Google exactly how the wrong URL came to be associated with the listings. My sense is that it often is incorrect “clustering” by the Local algo not necessarily an incorrect feed. I could be wrong about this.

    I recognize that Google thinks the greater community should somehow correct this error but I am of the opinion that if it is due to Google’s over aggressive algo than Google should get their algo together or perhaps be more cautious before they assign a URL.

    If the error is due to a feed, than I say perhaps a community edit is the way to approach it. However for the most part many of the “community members” are self interested, Google needs to do a fair bit of work making sure that the community is large enough and structured enough to make this happen correctly.

    Mike Blumenthal

  12. on 22 Jan 2009 at 6:38 pm admin

    Your comments are astute and to-the-point, as usual, Mike. I can’t help but agree with you regarding the importance of discovering not only where the problem occurred, but also, where the responsibility lies for correcting it. At present, this business is being sorely misrepresented.

    Mike, would you be able to shoot me a link to where you have written about clustering vs. feeds, if you have something. I’m sure I know what you mean by feeds, but am not sure I understand the difference between a cluster and, for instance, the old concept of neighborhoods. I’d like to get that straight.

    It is a great pleasure to have you comment here!
    Miriam

  13. on 23 Jan 2009 at 12:45 am Mike Blumenthal

    Information about clustering is spotty at best. Here is a bit about clustering from one of google’s locat patent applications:

    Structured and/or unstructured data about enterprises are acquired from one or more sources such as commercial data providers, enterprise web sites, and/or directory web sites. Strings are extracted from the unstructured data. The strings contain key, value pairs describing facts about the enterprises. The extracted strings are parsed to normalize the keys and values and place them in a machine-understandable structured representation. Some keys and/or values cannot be normalized. The facts are clustered with the enterprise to which they pertain. Normalized facts from different sources are compared and confidence levels and/or weights are assigned to the facts. These confidence levels and weights are used to select the facts that are displayed on a page for the enterprise in a directory.

    Another useful reference was this q&a with Google at SearchEngineland

    Q:: Can you tell us about authoritative sources in Google local? If a user makes a comment or requests a change, vs. a business owner, vs. a competitor, vs. a validated business owner, vs. a third party submission site (yellow pages, Yelp), whose content takes precedence? For us that is the biggest problem because it appears that Google takes that information and somehow creates a listing…not using the business owners listing.

    Google:: A business owner’s verified listing trumps all other sources in terms of fields displayed. LBC-verified listing is the most authoritative source. The least authoritative is a single reference on an unverified web page. Everything else is in between those two ends of the spectrum. In terms of creating the listing, we distinguish between the “listing” and the “cluster.” We display the “cluster” which is composed of the union of one or more listings. When the fields in a listing overlap, the listing with the highest authoritativeness trumps the others; but it doesn’t block additional fields (like cuisine, parking, etc.) from being associated with the cluster.

    Q:: Are you penalized for submitting your data on a weekly basis? Or should you let your data mature?

    Google:: No; however if you’re making changes to your listing that prevent us from recognizing that it’s the same business as the one referenced by other sources in the cluster, then there is a risk that your listing becomes “orphaned” from the cluster and thereby loses the associated content and any positive ranking from that content

    Mike

  14. on 23 Jan 2009 at 12:54 am Mike Blumenthal

    Sorry for the omission but the info about clustering in the local patent came from Bill Slawski’s article about Google’s Local Search Patent Application as well as the patent itself.

    Mike

  15. on 23 Jan 2009 at 2:49 pm Mike Blumenthal

    Interestingly “details” listed on Google Maps indicates that there are 6 websites that Google scraped or got a feed from:

    * Yahoo.com »‎ Has correct website url
    * Citysearch.com »‎Has correct website url
    * Chow.com »‎Has no url
    * Tastyr.com »‎Has correct website url
    * SuperPages.com »‎Has correct website url
    * Dine.com »‎Has no url

    So the likely sources for either a feed or a scrape all get it right.

    When I do a general purpose Google search on “Maya Palenque Restaurant” palenquegrill.com there are not other websites in Google’s index where they could have obtained the wrong URL data. So it doesn’t appear as though there is any internet source other than Google that has it wrong.

    That means that either an offline data supplier gave Google the wrong information or their clustering algo misinterpreted their scrape of the palenquegrill.com website and erroneously assigned it.

    My vote is on the later. If my analysis is correct (I have limited tools and understanding of Google Maps) then it is clearly Google’s mistake.

    Now Jen, why do you think the “community” should fix that?

    Mike Blumenthal

  16. on 23 Jan 2009 at 4:38 pm admin

    Wow, I am so thankful to you, Mike, for supplying such great resources here. I feel I am beginning to understand clustering after reading this. Let’s see….

    Basically, the cluster is an aggregate of facts Google has come across about the business, and while this should not, according to Google, trump the power of a verified LBC listing, it is definitely associated with the listing. This would be data from sources such as CitySearch, Yelp, YP, etc. and also from web pages that reference the business in some way.

    If you could not find any references linking Maya Palenque to the palenquegrill.com (neither could I, Mike)and it is only Google who is associating this business with this incorrect domain, your guess of a misinterpretation of the clustering algo’s scrape is a really good one. It sounds like a very smart guess to me, despite the fact that you have limited tools to investigate and understand this.

    I would like to second your question for Maps Guide Jen.

    Jen, should you stop by again, can you help us to understand why only Google seems to have gotten this wrong? If none of the 3rd party data providers has it wrong, why does Google?

    And…I’d like to add that I suppose there could be a possibility here of competitive hijacking as the listing has not been claimed. That’s a long shot, obviously. No one would really benefit from directing Maya Palenque’s customers to a restaurant in Texas, beyond simply sowing confusion. It doesn’t make sense that a local, competitive Mexican restaurant would benefit in any real way from such an action. But, because the listing is unclaimed, we always have to take into account the possibility of malicious edits, and the fact that the incorrect data is only apparent in one place (Google) would almost argue for the case that someone deliberately acted upon this single listing.

    While I don’t think that’s the case, it did occur to me.

    Thank you, again, Mike, for such a helpful set of comments.
    Miriam

  17. on 23 Jan 2009 at 6:58 pm Mike Blumenthal

    This would be data from sources such as CitySearch, Yelp, YP, etc. and also from web pages that reference the business in some way.

    Not to put too fine of point on this but also from the “enterprise website” ie the website for the business, school or agency in question. So for whatever reason in this case, Google seems to think that this is the website apparently based on the URL matching the business name…

    Jen?

    Mike

  18. on 23 Jan 2009 at 7:15 pm Mike Blumenthal

    Miriam,

    There is no indication the the record in Maps for Maya Palenque has been tampered with. There is no history of community edits so it was not hijacked via that route which in this case would be the only logical path.

    Mike

  19. on 23 Jan 2009 at 9:35 pm admin

    “Google seems to think that this is the website apparently based on the URL matching the business name”

    And that really doesn’t make sense, because the only similarity between the 2 things is the word ‘palenque’. To say that Maya Palenque and The Palenque Grill are similar enough to be confused with on another would be like saying Bob’s Burgers and Bob’s Seafood are the same thing.

    So, I don’t get this.

    Thanks for looking at the edit history, Mike. I didn’t think that was what is going on, but it did occur to me that a malicious edit would be ONE way something like this could happen to a business.

    Miriam

  20. on 24 Jan 2009 at 6:54 am Mike Blumenthal

    It doesn’t make sense, I agree. But Google’s algo, when working on the scale that it does (ie clustering data about the 15 million or so businesses in the US) seems to default to wanting to assign a URL. In the absence of constructive signals it seems to use very weak signals indeed to make that determination.

    The question is on a % basis how big of problem is it. Google always works on a % basis with these things. As much as you or I think that a business listing is sacrosanct, Google views it statistically.

    Given how many businesses there are and how many already have structured data available that provides the accurate detail of the url and of those that don’t have decent SEO on their site, how many mistakes like this does Google make?

    This has been characteristic behavior (assigning a URL to a record that doesn’t have one) of their Local algo as long as I have been writing about local.

    Mike

  21. on 25 Jan 2009 at 3:45 pm admin

    “In the absence of constructive signals it seems to use very weak signals indeed to make that determination.”

    I guess this is what we are seeing here. You’ve put that very clearly, Mike. And, it sounds like you’ve been seeing things like this for a few years now. I’ve only run into this a few times. I came across some very weird things with Western Union the other day that I want to go back a look at. Looks like it might be the same thing.

    Thanks for your great thoughts on this topic.
    Miriam

Trackback this Post | Feed on comments to this Post

Leave a Reply