Is Google Filtering Reviews or Reviewers?

stars on Google MapsA couple of weeks back I wrote a post outlining a review plan for local businesses. On that post I received a very thoughtful comment from David Mihm, a local SEO in Portland OR.

I was asserting that a business should get reviews from Citysearch and Insider Pages and Google almost as an after thought. David commented that he thought getting reviews on Google should be job #1. We had a friendly back and forth and than I read this post on Understanding Google Maps, which said that many of the reviews from Citysearch had been stripped out of Google’s results. With a resulting loss of “review stars” at the listing level and loss of business. Hmmmm… I thought, score one for David Mihm.

Applying What Was Learned From Links To Reviews? A Search for Quality

The question lingers, however, why would Google discount these reviews from Citysearch? I started thinking… if I were Google, what would I look for in a review? I couldn’t filter bad reviews and just show positive ones, obviously. No, I would want reviews from mavens. People who love to shop and give their well informed opinions. What would these people have in common? They would have a lot of reviews to their credit. So, I would look for reviewers with a large number of reviews to their credit.

Conversely, what if I wanted to game the system? The easiest way to do this would be to create an account, review the business and be done. Leaving in my wake a bunch of accounts with a single review on each, possibly 2 reviews to make it look good (but they were done on the same day). Is Google filtering these reviews? That’s a question worth asking… so I went to the florist that had lost some of it’s reviews and checked the reviewers.

Here are the Stats

Insiderpages shows 4 reviews and citysearch has 15. Google reports only the 4 from Insiderpages.

Of the 15 reviews from Citysearch (that google stripped out) 12 of them had been the sole review of the reviewer. Certainly, not mavens and possibly an attempt to game the system in the eyes of Google??? The other 3 reviewers had only 2 reviews to their credit… hmmm.

The 4 reviewers on Insiderpages had 4, 3, 2 & 5 reviews to their credit, better… but not earth shattering.

To look for more evidence, I went across country and found a florist in NY with only 6 reviews but the stars were given. When I dug into the reviews a little further, I saw that they were provided by reviewers with the following number of reviews to their credit: 83, 31, 69, 19, 55. Much more “mavenesque”

Moving on I came across a restaurant credited with reveiws from I don’t see that very often, as Google usually favors sites that specialize in dining reviews for restaurants, so I checked on the reviewers. The 4 who had reviewed that restaurant on had an astonishing 32, 191, 303 & 339 reviews to their credit.

Could reviews be getting like links? Where the number isn’t necessarily more important than the quality? Should the new strategy be to find people with a large of number reviews to their credit and give them a free sample? Are we going to have a paid review debate in the near future?

Some of these things probably aren’t happening yet but they may have just begun or be in the near future. As for Citysearch, most of the reviews on that site are from single reviewers. (Is Citysearch easier to game than others?) And that may have played a role in Google removing those reviews from the local algo.

And as far as the back and forth with David Mihm, I’m going to split the loaf and give David the bigger half: I would revise my advice to the small business owner to focus on getting reviews from Google and Insider pages and put a premium on those reviews from active reviewers.

This entry was posted in Reviews. Bookmark the permalink.

45 Responses to Is Google Filtering Reviews or Reviewers?

  1. David Mihm says:

    Tim, I think this may be the most insightful post I’ve ever read in the Local SEO arena. What tremendous research into and empircal evidence of the ‘maven’ idea — of course, as you explain it, it makes total sense!

    Instead of # of links = PageRank, it’s # of reviews = ReviewRank. Better to get 3-4 reviews with a lot of ReviewRank than 30 or 40 from someone without ANY ReviewRank.

    Paid reviews are already a big problem for Google and the other local engines; you’re absolutely right, though, that if this is true, the potential for ReviewSpam just got a whole lot bigger.

    Unlike with paid links, though, the difference is that the individual reviewers probably couldn’t care less if they lost the ability to pass “ReviewRank,” meaning it is a MUCH more difficult situation for the Google Local algorithm to deal with.

  2. Pingback: » The Most Brilliant Google Local Post I’ve Ever Read | SMB Web Design + SEO

  3. Matt McGee says:

    Tim – did you look at the age of the reviews that were purged?

  4. Thanks for the kind words David, it was actually your comment on the last post that got me thinking in this direction.

    Matt, I didn’t check the age of each of the reviews for this article but it might be interesting. It’s a good idea.

  5. Miriam Ellis says:

    This is a great post! Thanks to David for pointing me here.

    I can’t help wondering why Google has singled out CitySearch (apparently) as a poor resource. I also can’t help wondering if those accounts with 50+ reviews are legitimate, or clever Local SEOs who have figured this all out and are happily reviewing whatever they can in order to create power review accounts (think Digg etc.)

    Tim, I haven’t had the pleasure of talking to you before, but after this post, you can bet I’m adding you to my feedreader!

  6. Tim-

    Nice post and good theory.

    It appears that Google removed CitySearch reviews but also reviews from Google themselves.

    Certainly it makes sense if Google could actually develop a way of indentifying more trust worthy reviews and your idea might actually work…

    I will play the devil’s advocate. If Google were doing that (ie keeping reviews from more active reviewers) then we could assume that some CitySearch reviews would still be in their index. I haven’t poked around to look but that should be something we can ferret out.

    I also like Matt’s idea of time as a criteria. Certainly reviews become stale and worth less over time as businesses change. The bugaboo with that theory is that Google removed some their own reviews as well which are all very recent. But it would be interesting to look and see if the review dates are more recent than they were.

    To quote from an old Buffalo Springfield song: “Something is happening here, what it is ain’t exactly clear”…


  7. Mike, trying to ferret out a reviewer with a ton of reviews was actually difficult. The more I tried and failed the more I thought hmmmm. After several searches in several parts of the country I failed to find anyone with more than one or two reviews. I’m sure they exist, but the point is..
    If it is that hard to find one, is it easier for Google to just nix the whole thing?
    Does anyone know if Citysearch is easier to game than some of the other reveiw sites?
    And I’m sure age is a factor, interesting the way you put it – “Reveiws become stale.” It seems to me that Google gives more credence to older reviews. Which of course takes us back to Buffalo Springfield.

  8. Pingback: Understanding Google Maps & Yahoo Local Search » The Case of the Missing Google Reviews (cont’d)

  9. David Mihm says:

    Mike, Tim, et al:

    Related to the time theory, I’ve also found some evidence of a review sandbox on Google’s part (i.e. taking time before it “trusts” its own reviews). It’s for a client of mine, so unfortunately I can’t share the specifics, but let’s say in a market where the top-ranked, most-reviewed business has two G reviews from 2005, my client has 17 reviews from the last two months, a higher PR website, and is still appx. #8-9 in the Local algorithm.

    Obviously it might just take some time for Googlebot to find reviews on other sites like CitySearch, but surely it can update its algorithm to take into account its own reviews almost instantly?

  10. Michael Munz says:

    I agree with the fact that google has dropped alot of attention away from reviews. The one thing that was not mentioned was the overwhelming amount of negative reviews floating to the top of the serp’s……Is this a case of ” a few bad apples ruin te bunch”?

  11. Matt McGee says:

    @David re: review sandbox

    Google Maps as a whole is amazingly slow. They say it takes 4-6 weeks to just get a phone number updated, but the reality is you’re incredibly lucky if such a simple change happens within 2-3 months. If they can’t update their own data quickly, how can they update data from external sources quickly? :-)

    MB has done some pushing about how long it takes them to recognize reviews.

  12. Cathy says:


    Since you linked to my company’s reviews, let me enlighten you on the background of at least some of them.

    Our company has long believed our customers are our best form of advertising. With each emailed delivery confirmation (which we send automatically as a customer’s flowers are delivered locally) we include a request at the bottom of the page to rate us on one of the popular review sites – and include links directly to the pages. They’re linked from our home page, too. Some customers just email or write us thank you notes, but a few do go on to write reviews. I’m guessing most of them create logins to do so.

    At least a few of my fellow florists think we’re crazy and are begging for complaints. We’re not perfect, but we absolutely do pledge to keep our customers happy – and I hope that’s why we’ve earned their business and trust for the last 23 years. IMO requesting reviews is only a real ‘risk’ for companies who fear their customers.

    We were also the target of a black hat reviewer – a webmaster I ticked off for pointing out his/her spam. I blogged about the episode here: Turns out, that person created new identities and left defamatory reviews in two places. I was able to get the one from InsiderPages removed within 48 hours of finding it since I had the Florida IP of the supposed ‘local’ customer (we’re in California). Unfortunately though, Google Local picked up that review and it remained at the top of our Local profile for more than three months.

    You are probably right about a trust rank being assigned to reviewers based on the number of reviews. That certainly makes sense in the the case of the disappearing CitySearch reviews. (Only 2 or 3 of our 15 were ever displayed in our G Local profile.) Perhaps eliminating reviews – both negative and positive – from single reviewers may end up benefiting us all and encourage more users to say more, more often.

    With regard to the reviews left direct on Google Maps/Local, there must be more than a one-review filter. All but one of ours disappeared and the remaining one isn’t included in our count summary. I’m certain at least a couple of the reviewers had written about more than one company since I looked at their other ratings. :)

    What we should all be concerned about are reviewers like this one who use negative ratings to promote a particular website. Unfortunately for the reviewees, that spam at their expense hasn’t disappear. Go figure.


  13. Miriam Ellis says:

    Wow, Cathy, thanks for linking to that last profile. That pretty much says it all, doesn’t it? Incredible…and so stupid as it’s so obvious what the spammer is doing.

    David – that’s interesting that you are seeing a sort of sand-boxy thing, and I agree with the sentiment here that it takes A REALLY LONG TIME to get local data updated for clients when it changes.

    Tim – this has turned into a great conversation.

  14. Pingback: Understanding Google Maps & Yahoo Local Search » Google Maps Reviews- Was this review useful?

  15. I hadn’t noticed before this evening but google is now allowing users to rate the value of reviews. Clearly they are attempting to determine quality/trust.

  16. Will Scott says:

    This is a great post, and conversation. Thanks to David for pointing it out.

    The funny thing is that in responding to a post by Greg Sterling just this morning I was pointing out that there were a lot of theories but as yet no primer.

    I’m looking forward to some of the smart people in this thread validating your theory.

    We’ve got to make Google more responsive to the spammy BS going on in Maps.

  17. Will

    There are no “smart people” here only those who contribute to the conversation and this is all of us…by either asking questions, developing ideas, testing them etc….

    It is hard to write a primer on a “black box” that keeps changing but at least we can help each other understand as best we can its state today.

    To my way of thinking if google doesn’t resolve these issues (and if someone else does) they are risking their reputation and place in the local ecosystem.

    Small business people above all want fairness and equal access…if the system involves too much gaming they are going to give up on it before it is a real and useful factor.

    Mike B

  18. David Mihm says:

    “To my way of thinking if google doesn’t resolve these issues (and if someone else does) they are risking their reputation and place in the local ecosystem.”

    Mike, I’d like to agree with this statement…but with the incredible amount of search traffic Google receives day in and day out, I don’t think their reputation will suffer a whit, at least for several years. People seem to have this innate trust of Google, and although WE in the search community know there may be other, better options, G is still going to be the big dog for the foreseeable future amongst the general public, even in Local.

    I don’t think this will cause SMB’s to give up on local search, but I do think you’ll start to see more mainstream media stories highlighting Cathy’s experience as a local business getting spammed and having no one to turn to. I think if/when we start to see those, that’s when G will really start to invest some resources into clearing up these problems (i.e. lack of responsiveness to changes and updates, and lack of relevancy for certain reviews).

  19. Will Scott says:


    I hope I didn’t offend by suggesting you, Tim and the others responding here are smart :)

    You’re completely right in pointing out the critical mass issue. Users find valid answers or move on.

    That said, my personal thought-challenge is to take a truly empirical approach to the testing. It’s something I’ve been studying and wishing for an answer.

    As I understand the hypothesis: “fewer reviews from more trusted reviewers can beat more reviews with lower trust”.

    As it turns out, I know two providers of the same service in our home market (both of whom are family). I’ll document my proposed methodology, consistent with the hypothesis, and ping you (not necessarily) smart people.

    If anybody else reading also has 2 similar clients in one market it would be interesting to compare.

    Obviously there are pre-existing conditions, but if documented they may be accounted for.


  20. To Will:

    Hear, hear!!!

    I did not take offense at the “smart” reference but there is an element of self deprecation…the only way we can better understand the complexity of what we are dealing with is by each picking a small part, developing a rigorous research approach, gathering data and sharing results….together we are more competent and powerful than apart.


    I think from my reading at Google Maps for Business those businesses that have engaged are frustrated. The problems with the technology and the lack of transparency has pissed them off. I would venture to say that if you surveyed them that they have a lower opinion of google than the general population.

    The question for me is how much time does Google have to get its act together before this “attitude” rubs off…in some ways Local could be their Waterloo as it were.


  21. Will, I would certainly like to see the results of your test. I wonder if by the time you get it done, G will have changed again, though.

    More great reading on this on Mike’s post from yesterday… Getting the readers of reviews in the mix as another way of judging the quality of reveiws

  22. Pingback: Why Local Search Ruled the World Last Week » Small Business SEM

  23. Will Scott says:


    I tend to agree with you.

    Just look at InfoSpace and their initial local forays by way of YP publishers (going way back here — 1999).

    In some markets, the name InfoSpace still elicits growls and sneers from local businesses who bought the $4500.00 / month tile ad from their Yellow Pages rep.

    Bridge burned unequivocally.

    I do agree with David as well that Google’s halo (or horns) given their current transformative effect on small business online revenues may save their bacon from a few early missteps. The good news for Google is that true SME penetration (awareness even) is low once you get out of the top 20 or so MSAs.


    We’ll see how it shakes out. I think that the readers evaluating the reviews may have some impact, but ultimately it does have to be dealt with algorithmically.

    And, as I’ve thought about your hypothesis through the course of the day it does make sense algorithmically. I can generate 100 reviews for any business by morning. Just don’t pay any mind to my DP activity :) They’d likely be virgin reviewers and they’d be unlikely to appropriately geolocate.

    So, if we look at the algorithm of authority as “most trusted = hardest to game” then, a bunch of reviews from the Philippines for a US local service business are lowest in trust and a reviewer with 100 truly local reviews would be highest in this context.

    I promise, in my experiment the only Filipino reviewers will have just returned from trips abroad :)

  24. David Mihm says:

    @Mike: The frustration of SME’s with G Local is apparent, but I’m not sure how much that translates to the general public. I think there has to be a critical mass of CUSTOMERS frustrated with seeing spam / poor results before G loses its market position. And until more SME problems get picked up by traditional media, I just don’t see that happening. SME’s will be forced to do the best they can with G Local until their customers stop using it to find them :(

    @Will: Will, you make a great point about SME awareness outside the larger cities — the “long tail” of Local, as it were.

  25. Miriam Ellis says:

    Mike -
    I think your point is so important that the small business owners are frustrated. Clearly, the smart florists (like Cathy) who are finding the spam and know very well what it is and are justly angry about it must have about as low an opinion of Google as is possible.

    But, I think that’s behind-the-scenes turmoil, if you will. The average user (yes, I agree with the Google Halo theory put forth by David)just sees the finished product and as long as it is reasonably helpful, will probably be happy with that. Does the flower buyer really care whether the 800 number is truly going to a local business, or do they just care if their flowers get delivered?

    If consumers knew how small businesses are being spammed out of the results, they could object, but I think the problem is, most of them just don’t know the situation.

    In fact, the floral industry may be a particularly tough one because it has an element of mystery to it. Flowers arrive at the door…the recipient isn’t the purchaser…it’s all being handled by phone with the exception of walk-ins to the physical store.

    It’s different than the classic pizza case where you’re searching for a place 5 minutes from you to go eat. You’re going to be very annoyed and immediately aware if Google tells you to travel from California to Texas to get dinner.

    So enjoying reading everyone’s comments!

  26. Pingback: SEO Igloo Blog » Google Reviews Review

  27. Pingback: Trusted Reviewers Impact on Google Maps / OneBox | Website Promotion is not Voodoo

  28. Matt Ross says:

    Very interesting…chalk this message up as one more review towards my future Google review of relevance.

  29. Pingback: SEO Igloo Blog » Google Maps Not Counting Own Reviews

  30. Very insightful. So good infact – it might have been better to keep it to yourself ;)

  31. Pingback: » @Search Engines: How to Improve Your Local Search Interface | SMB Web Design + SEO

  32. Pingback: SEO Igloo Blog » Local Hound Tim of Convert Offline - Local SEO Interview 3

  33. Pingback: location jeux casino

  34. name says:

    Really nice=)but look this:,

  35. Pingback: Loci 2008 - Will Scott’s Best of 2008 in Local » Understanding Google Maps & Yahoo Local Search

  36. Pingback: Will Scott Featured in Mike Blumenthal’s Loci 2008 | Website Promotion is Not Voodoo

  37. Rimma says:

    Just do it: ,

  38. Kelvin47 says:

    Yet we have not got it convincingly or steadily right. ,

  39. Pingback: Google Maps: Citations, Reviews, & User Content

  40. I guess google is not filtering anyone

  41. evillersiog says:

    lone worker
    lone worker policies
    lone worker policy
    lone worker alarms
    lone worker alarm
    lone worker safety
    lone worker protection
    lone worker risk assessment
    lone worker systems
    nhs lone worker policy
    nhs lone worker

    The Health and Safety Executive (HSE) is frequently asked these questions. Working alone is not in itself against the law, and it will often be safe to do so. However, the law requires employers and others to think about and deal with any health and safety risks before people should be allowed to work alone.
    Employers have responsibility for the health, safety and welfare at work of all of their employees. They are also responsible for the health and safety of those affected by work activities, for example any self-employed people they engage and visitors such as contractors.
    These responsibilities cannot be transferred to any other person, including those people who work alone. It is the employer’s duty to assess risks to lone workers and take steps to avoid or control risks where necessary.
    Employees have responsibilities to take reasonable care of themselves and other people affected by their work activities and to co-operate with their employers in meeting their legal obligations.

    lone worker monitoring
    lone worker training
    lone worker solution
    lone worker solutions
    hse lone worker
    lone worker legislation
    lone worker solutions
    lone worker solutions
    lone worker gps
    lone worker risk assessments
    lone worker device
    lone worker risk assessments
    hse lone worker policy
    lone worker procedure
    argyll lone worker
    identicom lone worker
    identicom lone worker
    lone worker alarm system
    lone worker panic alarm
    identicom lone worker
    lone worker regulations
    lone worker definition
    lone worker alarm systems
    lone worker software
    lone worker safety training
    lone worker software
    lone worker risk assessment template
    lone worker risk assessment template
    lone worker risk assessment template
    lone worker risk assessment template
    guardian 24 lone worker
    lone worker safety solutions
    guardian 24 lone worker

  42. Pingback: Mihmorandum | “Putting” Google’s Citation Algo to the Test | Google

  43. Pingback: Using Third Parties to get your customer reviews | Successful Internet Tools

  44. Pingback: Best Old Posts on Local Search: the Classics |

  45. Chris Powell says:

    I have had some clients complain about reviews being removed and in most cases they tend to be older ones. I was just looking at one case where a review I had posted had disappeared and I speculated it was because I had not left a review in a while. I logged onto Google and left three new reviews and I am speculating that the old review may return. In this case, my review is from two years ago and all the existing reviews are from within the last year. I am going to keep an eye on this and I will try to check back in with an update to let you know if the old review reposts. Another issue was that Google also required me to authorize which reviews I was sharing with whom. I approved most reviews to be shared. We will see how this works into it as well.