Google Places Troubleshooting: Best Practice for Dealing with a Merged Listing

Google’s problem solving paths for fixing things that go wrong with a business listing have always been confusing. There are a range of possible options (Report a Problem, MapMaker, Troubleshooters, Places Dashboard) to fixing a problem  and there is the temptation to keep trying one after the other in an effort to get the problem fixed. Most SMBs and SEOs figure that getting a simple data error fixed should happen in near real time and think that if they just tweak one more variable then the fix should be in tomorrow.

The reality with Google Places is that doing more and trying more and not being patient will lead to more, overlapping problems as Google reconciles the many data sources that are affected by your repair efforts. This is true in most cases but particularly true with the most serious & deleterious artifact of their algo, the merging of two distinct business listings. Reader Susie made this recent inquiry:

It seems that Google Places is a fan of combining businesses.. We have been having a bit of trouble with our listing lately. Google places has somehow combined our business with another competitor’s… Our website and placement on google places will show up, but underneath our website our competitor’s phone number will appear. Keeping in mind that our addresses are the same (as it is a dock location) and we do have similar competing services………… How do we insure that our business phone number stays attatched to our website in the google places format and not our competitor’s? We have already tried switching keywords, posting pictures in order to differentiate our business, but that has not seemed to be working……….. Any suggestions would be greatly appreciated.

Nothing can be more frustrating and nothing needs a more Zen like approach than merged listings. Here is my advice on the best way to currently fix the situation:

@Susie

Step 1) Make sure that your name is short, that it is your actual legal name and different than your competitors. Change the Places listing to reflect this name. A long name, particularly one with the same keywords as your competitor can contribute significantly to the problem. This is critical long haul as your business name is one of the primary items that you have control over that causes merges and it is an important signal to keep your cluster intact.

Step 2) The only sure way right now to get a merge taken care of is to have Google split the two listings. Once it has occurred it is almost impossible to fix by any fiddling you might do with your data at Google. The recently added Fix a Problem troubleshooters on the Google Help page offers the best current option for a fix as it goes directly to a well trained staff that have the tools and the knowledge to solve the problem. Use the My listing has incorrect information option and take  the following path on the questions:

-What’s the status of your listing?
—I have verified my listing in my Google Places for business account.
-Have you tried the Report a problem link?
—Yes (or it is not available on my listing)
- If you have attempted to update your business’s information from within your Places for business account and have not seen your changes take effect within a few days, you can complete this form to specify your correct business information.

Step 3) Once you have filed your repair request with Google, do not keep changing things or trying to fix them with one of the other data input or repair options. Google will respond within about 10 days and it will take 4-6 weeks for the fix to become “permanent” and public. In the meantime STOP play with it. Be patient and wait. This is the most difficult part to understand and by far the most difficult to do. Google assembles the data for the Places index in a batch way, collating and reconciling changes over a given period before pushing out the new data. The Troubleshooter team has the ability to inject correctly unmerged data into this collation but it is necessary (hard to believe yes but necessary) to wait for the corrected data to show up in the next index push.

Step 4) You need to be sure that going forward that the signals coming from across the local ecosystem to Google’s algo send clear, consistent and distinct information about your business listing. Verify that your business name is listed properly with the primary list providers to Google in your country, get listed identically on every directory, local resource and travel guide that you can. Be absolutely sure that you use the legal business name (DBA) noted above. Clean up any errant records that might have a variation in name, address or phone number. Going forward, continue in your efforts to obtain these mentions of your business at your location as they will reinforce the view that Google has of your business as distinct.

Good luck and may the gods of fortune and patience be with you.

Please consider leaving a comment as your input will help me (& everyone else) better understand and learn about local.
Google Places Troubleshooting: Best Practice for Dealing with a Merged Listing by

26 thoughts on “Google Places Troubleshooting: Best Practice for Dealing with a Merged Listing”

  1. Thanks! Just sent this to a small business friend I know – he’s working on fixing the merged listings himself and I know it’s been frustrating for him.

  2. Thanks for the heads up! I had a question for you though, and you’re the local guru so I figured you could help.

    Recently, I had a business listing of mine ranking well – it was a blended result with a solid Places page. However, mid-December the Places listing was dropped from the result (though I’m still ranking very well, organically) and the Places page won’t show up within a Maps search either.

    The only way I can find my Places page is to login to my dashboard and access it that way – and it shows up fine, is owner-verified, and is active. In fact, Google is reporting (as of yesterday) 35 impressions! This can only mean that its showing up somewhere, right?

  3. Oh Mike that summed it up so well! Merges are so frustrating!

    I totally get it that they would keep tweaking this and tweaking that thinking they can fix it. Just like troubleshooting your computer, you just keep trying different options til you find the problem. But like you say that’s the worse thing they can do and just compounds the problem.

    I was going to link to a bad merge problem where the atty DID try to go through GP channels to get his merge fixed and that was a nightmare too. But I see you already blogged it here: http://blumenthals.com/blog/2012/01/19/the-long-and-winding-tale-of-trying-to-fix-a-merged-listing/

    We see these merge problems every day in the forum and to me it seems like lately it’s getting worse. Would you agree?

  4. @dana

    Great. Be sure to let me know the outcome although I am sure that he has already clicked every possible option which will literally pollute the process as you can’t tell where communication is coming from and which section of Google is doing what to your record. Since these other methods all have their own time to fix it is near impossible once you bombard Google to understand any of the communications coming back.

    @Curtis
    There are several possibilities for the outcome and you haven’t really left me with enough information to know which is likely.

    @Linda
    Yes, I am seeing more merges both within multi practitioner businesses AND amongst different businesses. This effectively means that Google has turned up the knob on the dupe purge algo… it will result in fewer dupes but more merges.

  5. This client and his wife run two separate businesses with very similar names out of the same house – I suspect that a Yellow Pages import originally messed up their listings and they just couldn’t get them separated. At the time, Google support was useless. I hope he can get it fixed now!

  6. @Dana

    Hmm… this could be the first case in recorded history that the outcome of Google’s dupe purge routine has lead to a divorce. PLEASE KEEP ME POSTED. That would be a great story to break. :)

  7. Mike – NY Times, here we come! Google Places Merge Leads to Divorce! Good thing we know the reporter there so we can break the story when it happens. ;-)

  8. @Mike,
    My sense is that there are both more purges and more duplicates. And both seem more “persistent” than they were even a few months ago.

  9. @Jim

    All of these observations (more dupes, more purges, more merges, more persistent) could be explained if Google introduced a new trusted source feed into the data mix. And if it were necessary to crank up the merge/purge routine to deal with the increased influx of this trusted that was creating a number of dupes…. Hmm this is beginning to sound like MapMaker to me. Your thoughts?

  10. My first thought is tangential. The ‘Location not supported’ errors we have been seeing over the last few weeks seem to me to be related to marginal violations of the quality guidelines. My sense is that is a symptom of a ‘cranked up’ spam algorithm. Nothing really to base that on, just a hunch; but I’ve learned to trust my data hunches.

    The dupes really have me baffled. So many of them seem to be identical NAPs; many times the second listing only has one Category. Why two data clusters, CIDs, for one NAP? A simple (one category) one coming from another source and integrating poorly with the existing cluster in Maps?

    I wish I had some sense of how MapMaker works. I’ll start reading with the possibility in mind that MapMaker is becoming and independent data source from the regular local data cluster; that assumes that MapMaker and the local data cluster were the same, or at least more integrated, in the past.

    My sense is that there are 3 data sources for Local Search. Place Pages, Reviews, and the local data cluster. The local data cluster is a multi-source database of its own consisting of the data scraped from the internet and the UGC of MapMaker.

    Hm, you’ve got me thinking anyway.

  11. Jim

    Your thinking about where and how Google handles incoming data for Places Business data is essentially correct but too constrained. When you read the patent language you realize that they source data as far and wide as they can and that it is handled differently depending on structure and the differing levels of trust. Not all data is scraped, some comes in via feed or is purchased and fed into the system.

    My assumption about MapMaker is that it has always been considered a separate data source but that what has changed is the level of trust. Remember it was first used to map 3rd world data and only after it was used for quite some time did that data finally come into Maps. If I recall Kenya was the first data set to be rolled into Maps. Then Mapmaker was gradually rolled out across the 3rd world and was finally rolled out in the US. It has only been recently released into Canada. And is still not available in much of Europe.

    Google loves to triangulate data from multiple sources. They love cheap data. They also think that they can gather and create good quality data better than anyone else.

    Google MapMaker is a way to get very inexpensive crowd sourced data from a very cheap source. This allows Google long haul to get data in countries where the primary data sources don’t or won’t share AND once the product gets enough data allows them to get rid of data sources in the US that they have been paying for.

    This pattern has been quite clear with underlying Mapping data vis a vis Navteq and TeleAtlas. First they buy the data, then they replicate it (Streetview), then, once it is “good enough” they integrate their data in and create a crowdsourced solution to correct it.

    View MapMaker business listing data in that light.

  12. Re the original issue:

    Obviously without seeing the Places entry data we cannot tell where these are from.

    In Mapmaker and merged into the Places listing are these spam Categories:

    St. Petersburg Drug Crimes Attorney / Lawyer, St. Petersburg Juvenile Crimes Attorney / Lawyer, St. Petersburg Domestic Violence Attorney / Lawyer, St. Petersburg DUI Attorney / Lawyer

    These need to me corrected in Mapmaker and any other web entry.

    This Share an update has over capitalisation spam:

    Melinda Morris, former State Prosecutor, is ready to take your call and answer your criminal defense questions. –> MORRIS LAW FIRM – CALL 727-388-4736

    Also do the rules that apply to Descriptions apply here as well?

    In the description:

    DUI & Criminal Defense Attorney, St. Petersburg, Florida. Experienced former State Prosecutor who handles DUI, Domestic Violence / Battery, Juvenile Crimes, Drug Crimes, White Collar Crimes, Traffic.

    This is pure location and category spam, not a description.

    Any one of these can stop the Places entry sourced data getting into a blended Maps search result and in that situation your Maps index cluster will be merged with a nearby business. You can stop this sort of merge by making the Places entry, and I assume the related Mapmaker entry, compliant with the implemented spam checking.

    Cheers. Andrew.

  13. I hadn’t seen that patent before, and it’s interesting. Google went in in 2005 with very broad claims, and five years later, after much fighting with the USPTO, got a patent (#7,788,293) that only covers extracting the business hours of a company from its web site. The patent examiner didn’t think much of Google’s original claims. (You can read the whole history of the examination by looking up the patent in the USPTO’s “Public PAIR” system, which has all the correspondence back and forth. It’s all a public record.)

    Google’s data sources seem to be quite limited. This becomes clear from their mistakes. For a while, Google’s own place page listed them as a solar energy business. I pointed that out on Search Engine Watch, and within a day, they’d claimed their own place page. Google Places tends to have bogus data for large companies. Ford Motor corporate HQ, for example, is listed as “Medical Clinic”. It’s even worse for B2B companies. Mosler Safe Company is listed as “Construction”. Caterpillar is listed as “Fire Department, Insurance Agency, Tractor Dealer”. Applied Materials, the world leader in semiconductor fab gear, is listed as “Business, Medical Supply Store”. It’s clear that Google isn’t using any data source that has real information about companies. It looks like they’re still using “throw a classifier at it and hope” technology.

    In the consumer area, Google Places thinks Coit Tower (a San Francisco landmark) is a carpet cleaning service.

    After years of crappy Places data, companies screaming about it, and embarassing criticism by both the New York Times and Fox News, you’d think they’d be using better technology. (Like ours, but I digress.)

  14. @John

    With the size of Google’s local data set it is difficult to know if the egregious errors your point out are statistically significant or not. Google would contend that even the best list that they could buy would have a 15% error rate and that they have managed to improve that.

    Every list has outdated listings that have closed and have not been removed (false positives- real listing, show open but they are not) but Google’s methods, in attempting to create a lower error rate in the list create what I would call false negatives (listings that don’t really exist but show as open) in the form of merges etc.

    The questions are
    1)Is their list in aggregate any better statistically? If so by how much?
    2)Are the false positives worth the high social cost?

  15. Lets’s try the biggest companies on the Fortune 1000:

    Wal-Mart Stores Headquarters – Google Places says “Pharmacy”. The second search result is Best Buy headquarters – FAIL.

    Exxon Mobil Headquarters – Google Places says “Service Station”. Location correct. No totally bogus data visible. – OK.

    Chevron Headquarters – Google Places says “Service Station”. Location correct. Photo is of some clothing hanging in a closet. Associated ad is “Shape Any Look With Ease. Check Out Gillette® ProGlide Styler Today.” – FAIL.

    ConocoPhillips Headquarters – Google Places says “Service Station”. Location is a ConocoPhillips warehouse in Savannah, GA. The only result even close to the right location in Texas is a motel that’s near their HQ. – FAIL

    Fannie Mae headquarters – Google Places says “Bear Stearns Office
    383 Madison Avenue, New York, NY. This place is permanently closed.” Yes, Bear Stearns went bust, but Fannie Mae lives on, propped up by bailouts. – FAIL

    General Electric headquarters – Google Places says “Ge’montes Headquarters Barber”, in Anderson, SC. – FAIL

    Berkshire Hathaway headquarters – Google Places says “Lodging, Marketing Consultant”. That’s Warren Buffett’s company, and they own other companies from Acme Brick to XTRA space. No lodging companies. Their main business is is insurance. Google Places at least got the address right. So – FAIL

    (If you try these searches in Google Places without the word “headquarters”, Google tends to return some random company location. “General Electric”, for example, returns GE Lighting in Wayne, PA. Category is “Lighting store”, which isn’t a total miss. Fannie Mae it gets right, as “Mortgage Lender”. They don’t have branches, so Google doesn’t get confused.)

    That’s pathetic performance. Those are the biggest companies in the world and Google gets most of them wrong.

  16. @John
    Anecdotes are common in the world of Places. And the failures abound BUT that still doesn’t tell us anything that we didn’t know…. which is that some listings are not accurate.

    The questions are
    1)Is their list in aggregate any better statistically? If so by how much?
    2)Are the false positives worth the high social cost?

    I assume you are saying that even if the list is better that it is not worth the social cost. I still, after all of these years, withhold judgement as to whether their experiment in big local data is a success. The reasons being that while it fails for some I have seen it work very well for many a small businesses, I don’t really know how big the problem is and I (again anectdotally) have seen it improve since the very days…

    Mike

    PS I am not sure that your search methodology is a very good way to ascertain accuracy… for example when I search on Walmart Headquarters I get this answer (which is pretty good):

    Best guess for Walmart Headquarters is Bentonville, Arkansas, USA
    Mentioned on at least 4 websites including clui.org, about.com and hoovers.com -

  17. That’s the web search engine, not Places. Wal-Mart HQ is in Google Places more than once, and where you’re searching from can affect the results. Try “Wal-Mart Stores Headquarters Bentonville AK” in Google Maps (“Places”). That gets you the entry tagged as a “pharmacy”. If you try it without “Stores”, you get a motel near their HQ.

    Go down the Fortune 1000 list. See how few big businesses Google gets right.

    Real data on businesses and where they are is available. Google doesn’t use it. That’s how you were able to get “Illusory Laptop Repair” into Google Places. It’s still in Google Places, too.

    Bing could eat Google’s lunch in this area if they ever figured this out.

  18. @John

    That just tells me that the Fortune 1000 hasn’t done ANY local SEO and they they have’t bothered to watch over the quality of their NAP. There are 20 million businesses in the US.

    As you know I have no love lost on Google but I would ask you who is doing a better job of getting local data right at a worldwide scale?

    With Google the issue is that you see the bad data. That doesn’t mean that it is worse. This comes back to my original question of “Is their list in aggregate any better statistically & if so by how much?”

    Anecdotes might point to in the direction of a question but they can’t provide an answer

  19. @Mike Quite agree with you their Mike. Google is doing a better job than ‘The rest’ when it comes to Local search. I think that these Google local listings are the places to go for on the front page in the future. It’s Googles front page, and it looks like they want it for themselves…

  20. I have all my clients under my Google places and am wondering if that is a bad idea. Does that hurt their Google places optimization?

  21. Cathie, that’s a really bad practice for other reasons.

    1) If you part ways with client you can’t give them log in to their own existing listing.

    2) All it takes is violations on 1 or 2 listings and G can suspend your ENTIRE account, which hurts all your clients, even the ones that don’t have violations. We have SEO and design companies come to the forum all the time saying: help my account with 50 or 200 clients just got suspended. It’s a NIGHTMARE!

  22. Hi Linda,

    Where SEOs have set up a Places entry in a free non-gmail email account, my advice, when Places and ‘Google+ Business’ are merged will google make a new gmail and Name+ profile with a ‘Local’ Business page in there?

    Cheers. Andrew.

  23. So, I followed the procedure mentioned above. I just performed step 1 and 2.

    After filling the very short support form out I get a message on the screen: “Thank you for contacting us.”

    I thought that Google would’ve at least give a text box so that I could explain to them that an attorney listing was merged with the main office listing.

    I thought that I would need to show them both of the URLs so that they can ‘unmerge’ the 2 listings… but they don’t allow you to add comments to the support request.

    Well… I shall move on to step 3 and 4… the waiting for 10 days… though step 4 will mostly likely be an ongoing monitoring of the major data aggregators + their leeches… errr… ‘partner websites’ :)

    Russ

  24. I called my Google Adwords representative and complained that I could not buy Google places ads because of the merge and it was fixed in 3 or 4 days which surprised me.

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.