|Google Suggest - Can We Affect the Negative Entries?|
| 3:01 pm on Oct 4, 2009 (gmt 0)|
Google Suggest results are hurting my business. When typing in my domain (which people do for whatever reason in a search engine) it shows "MyDomain.com" as the 1st result but shows my domain with negative words about my domain as the 2nd and final suggestion "MyDomain.com sucks". Im needing to alter the Google suggestions. I saw a competitor of mine had this "sucks" term listed on their site at one point but am not sure if thats the reason its now suggested and connected to my domain. I doubt many people would actually search for "www.mydomain.com_sucks". Its all a bit odd.
If I knew the exact source that Google references to post the suggestions, it would be a great start. Google Trends shows that there are not enough searches with the term (the one including 'sucks') to have any data to display. My questions are as follows:
If I manually do searches on what I want to show up (100 daily for a month) will it later be a "suggested" term offered by Google Suggest (assuming the collective world users do "mydomain.com_sucks less than that many times in a given period)? Again, the suggestion I want removed is my full domain url with a negative word which is not searched for often if ever. Yet, everyone now sees it when doing a search on my domain.
Does Google Suggest "only" use the amount of searches from other users to determine the results or does Adwords play a role for their Google Suggest suggestions? If Adwords is considered, couldnt I just add those words to my keywords in Adwords to alter the suggestion results?
How often are the trends (or whatever data source they use) updated for Google Suggest? In other words, if there were 100 searches in one recent week for one keyword combination but 500 collectively over a 5 year period, would the 500 old results out weigh the 100 new searches in determining what is suggested in Google Search? And how long would it take for all of the new trends to be accounted for in Google Suggest?
Google Suggest states they use niche audiences for what is suggested. Australia may be different than what the U.S. suggests. Is this based on the actual IP address of the person doing the search or is this based on the various google sites (.com, .com.au, .co.uk, etc....)
5- I have no doubt that the suggested results is due to my competitor since not too many people put a domain in a search engine and then the word sucks. My competitor did have this on his site though at one point and could have possibly added it to his adwords at one point. So, if anyone knows a way to simply find out the exact methods used for the suggestion tool (that which is weighed in giving the results such as current listings, adwords, other's searches, words from top websites, etc.....) then that explanation would be appreciated.
By the way, WebMasterWorld is the absolute best resource for any question. I have referenced it often and have always found what Im looking for. Real world webmasters with real world solutions. I applaud all of you who participate!
| 5:57 pm on Oct 4, 2009 (gmt 0)|
Hi MookiesMatter, and welcome to WebmasterWorld.
|Google Trends shows that there are not enough searches with the term (the one including 'sucks') to have any data to display. |
Several comments... Google Trends is useful only for searches with lots of volume. It is not what I'd call a very "granular" tool otherwise. So it's extremely unlikely that Trends would show searches for your particular domain, and even less likely for a more specific search combining your domain name and an adjective.
|Does Google Suggest "only" use the amount of searches from other users to determine the results or does Adwords play a role for their Google Suggest suggestions? |
Suggest was a Labs project before it became a standard feature on the search box. In the Labs incarnation and in the early days of the search box implementation, numbers were returned along with the phrases. The numbers then reflected the number of pages returned, not search popularity.
Ultimately, Google dropped the numbers, and I think the official word (or popular supposition) was that they did so because they were distracting and people weren't really looking at them. But it's also likely that Google dropped these because they had factored other considerations into the personalized (and perhaps other) configurations of the tool, that the numbers therefore weren't really correlating any more with what the tool was showing in the personalized setup, and it was too complicated to switch displays.
Google's Web Search Help says that historical data is used only when you're logged in, so it may be that you're seeing what you're seeing because you are logged in. In the More about Google Suggest [google.com] section of the Help page, Google describes it this way...
|As you type, Google Suggest communicates with Google and comes back with the suggestions we show. If you're signed in to your Google Account and have Web History enabled, suggestions are drawn from searches you've done, searches done by users all over the world, sites in our search index, and ads in our advertising network. If you're not signed in to your Google Account, no history-based suggestions are displayed. |
I don't know if you're seeing the problematic suggestions just when you're signed in, or all the time.
Also, if you are signed in... it's not clear from the Google Help description how much your own historical searches are affecting the results you're seeing, vs historical searches from everyone else.
It may be that you're skewing the results just for yourself, or that if you're searching too much for "MyDomain.com sucks", skewing the results for everyone. You'd have to sign out, flush cookies, and sign in under a different account to observe differences.
As an aside to this, there is currently in the Supporters Forum a discussion about what features a useful autosuggest tool should have....
WebmasterWorld Supporters only...
The Importance Of Autosuggest For Search & Ecommerce
| 8:09 pm on Oct 4, 2009 (gmt 0)|
Wow.. the Labs project ey? You are very informative and this is quite interesting. I made sure i was not logged into any certain account when doing the google search. I tried with four browsers with no cache as well as had others in different states try it to confirm my findings. I did the same search with other computers I have that have never been on the internet just to make sure it wasnt a cache thing.
Later, I tried accessing it with a proxy server and various IPs connected to different regions but the same results. What is unfortunate is that there is just my domain and the tainted domain suggestion. So it doesnt even give the 10 or so other suggestions. Just the two.
You make a good point about me skewing my own results by my own input of the information to see the results. So I therefore did tons of searches with my domain and other various words "besides sucks" to see if they would catch on to the suggestion. I used different IPs, browers, and no cache each time. It seems that the only two original results remain still. Those searches didnt even add my multiple other word searches to fill in the next 3rd, 4th, etc.. results. 8 of the top 10 suggestions are still blank.
I appreciate your input! Ill look into the Labs project more and see if that helps. Many thanks for your time!
| 11:33 pm on Oct 4, 2009 (gmt 0)|
|...just to make sure it wasnt a cache thing. |
Just so there's no ambiguity, cache and cookies are different, and browsing histories would be stored by Google and identified either via cookies or by account sign in. Since you're not signed in and cookies in different browsers are stored in different places, and you tried four of them, it's not likely that this is personalized data you're seeing.
But this is a reputation management kind of issue, analogous to but not exactly the same as having an unfriendly site in the top ten (or twenty) that ranks for [MyDomain.com sucks]. In serp-related reputation management, you need to drive the offending phrase down by creating pages with other phrases that will be commonly associated with "MyDomain.com".
With Google Suggest, you're not necessarily dealing with phrases that rank in the top 10 for MyDomain.com. You're dealing with commonly occurring phrases of which MyDomain.com is one word. Though right now, "MyDomain.com" and "MyDomain.com sucks" are the only 2 suggestions appearing in Suggest, the Suggest list will expand to as many as 10, so you're going to have create at least 9 other combinations associated with your domain name and tld that outweigh any offending phrases. This is not easy.
While we can't discuss specific domain names or search terms here, we can mention, say "microsoft.com". Type it in the Google search box and watch how the suggestions change as you type each letter... and note what happens when you get to the dot in .com, and then the "c". Finally, you'll end up with a set of products associated with "Microsoft.com" and subdirectories on the Microsoft site, which are also of course associated with Microsoft.com. You're going to have to find a way to do something similar for your domain.
Also, look at Google's suggestions for various competing domains, and note what pops up as you type their names. You'll see that there's a fairly common set of negatives that appear for domain names in any competitive market area.
Current law protects the various "review" sites that put this material up. They represent the spectrum from free speech to extortion rackets of sorts. Many of them survive by selling ads.
Note that you'll need to use Google search find out how many such pages for each phrase combination you're dealing with to get "mydomain.com sucks" and other negatives that might occur off the list.
|I'll look into the Labs project more and see if that helps. |
Again, Google Suggest was a Lab project. Now it's a standard feature.
| 7:53 am on Oct 5, 2009 (gmt 0)|
Robert, thanks for the response. Getting results as Microsoft did would be great but I am wondering... do you think those subdirectory results such as /security, /office, /hardware etc... were in the suggested results because people actually have them memorized and did a search for exactly that directory? I see that one of them is /office/2007 after the .com which makes me wonder if the "suggested" results are significantly altered by something other than that entire address being entered into the search engine. I know those results have tons of links linked to them throughout the web and wonder if there is more emphasis on that link popularity opposed to the keyword search. If this is the case, it would be easier to optimize some back links in my case.
Since I have 8 blank suggestion spaces to work with, it sounds like I would be able to fill in some of those suggestions fairly easy since there is no competition. Any idea how often Google actually updates this type of information to reference their results?
Your comments are truly appreciated!
| 10:31 am on Oct 5, 2009 (gmt 0)|
|Ultimately, Google dropped the numbers... |
I don't think they dropped the numbers, I'm still seeing them.
| 11:07 am on Oct 5, 2009 (gmt 0)|
The suggest numbers (results) appear at my end too
| 1:09 pm on Oct 5, 2009 (gmt 0)|
Fascinating that some still see the numbers next to the Suggestions. I haven't for quite a while. Apparently it's still in testing of some kind and the decision isn't "final".
| 9:04 pm on Oct 5, 2009 (gmt 0)|
|...do you think those subdirectory results such as /security, /office, /hardware etc... were in the suggested results because people actually have them memorized and did a search for exactly that directory? |
No. As I'd noted, when Suggest was a Lab project the Google Suggest numbers originally correlated roughly with the the number of pages returned that contained suggested target phrases, not with search popularity.
Google's help page indicates that if you were looking at personalized results, Google would also be factoring in search popularity. But, if you're not signed in, "no history-based suggestions are displayed." This latter phrase, "no history-based suggestions", of course, might be intentionally ambiguous in terms of how search history enters into the Google algorithm as a whole.
As it turns out, Google is the only engine that returns subdirectory suggestions for microsoft.com. I see Google suggesting a very occasional subdirectory for other domain searches, but they're not nearly as frequent as they are on Microsoft. So these may be very special cases... perhaps worth exploring.
It's very difficult, though, to get straightforward search results (on any of the engines, not just on Google) for searches including the slash character, so this subdirectory co-occurrence is not easily described by simply saying that if the terms co-occur enough on a page, Google will include them in Suggest. Nor does treating the slash characters as delimitors give a clear picture of what Google is looking at.
It may be that Google is coming up with its subdirectory suggestions by looking at a combination of domain.tld co-occurrence with both the keywords in the urls and in occurrences of the full subdirectory urls on the pages. (Microsoft urls often appear in full on referring web pages). Google might also be using weighting factors analogous to those they use in choosing Sitelinks.
Google is also using phrase-based indexing at this point, not just word matches, and this further complicates simple considerations of number of pages returned for associations with your brand name.
But, bottom line... you've got to do a lot of online "promotion" to get your brandname associated on other websites with terms other than the word "sucks".
| 10:52 pm on Oct 5, 2009 (gmt 0)|
Okay, I think I will try a few things to see what works best.
1- I am going to add "mydomain.com" plus a random word in adwords to see if it eventually catches on in the suggested results (with Google Suggest that is). I understand some sponsored results could possibly show up such as a search for AOL's sponsored ad when searching for "aol". It would be simple to just run the ads I want to show up pending any restrictions... I did this a few days ago but am still waiting to see if any results change once they update whatever database they use to determine these suggest results.
2- I will also create a subdirectory to see if I can duplicate Microsoft results with another term and get that /subfolder indexed.
3- Finally, I am going to do what Google did. It appears their results will have co.uk, com.au, com.vn, etc... that show up with searching on their own domain.
Hopefully one of these methods will work.
What I am trying to avoid is optimizing tons of "mydomain.com sucks" pages to be brought up with the google results. I would hate to find out later that my optimization of "sucks" is actually a reason it is showing up in the "Google Suggest" results.
If anyone has an idea or is in the same situation, please let me know and I will try other methods that may be helpful. As previously mentioned, only 2 of the 10 "Google Suggestion" results bring up any information. This means I am not really competing against a huge number of searches across the world to influence the 8 extra slots. I may be in a situation where this is actually fixable. I will post what I find works the best when I get something showing up.
Robert, I know you are correct when you state "But, bottom line... you've got to do a lot of online "promotion" to get your brandname associated on other websites with terms other than the word "sucks". Do you think the three methods I laid out are viable options (and possibly less time consuming) than doing a full brand name association campaign? I most definitely will be doing the brand name campaign but would like to find a quicker alternative (for right now) to void the Google Suggest results immediately until I can pursue the more time demanding alternatives.
| 4:38 am on Oct 6, 2009 (gmt 0)|
|Do you think the three methods I laid out are viable options (and possibly less time consuming) than doing a full brand name association campaign? |
No. I think you're assuming that getting Google to show suggested searches together with your domain name is going to be a lot easier than it is. Though, as you mentioned, there's no competition, there is definitely a fairly high threshold for suggestions to appear in the Google suggestion box.
If you look at most domain names that appear in Suggest, you'll see that not very many of them have accompanying terms as additional suggestions. I'm sure there are weighting factors of various sorts for the suggestions that do appear. For many domains, the suggestions offered are quite often the result of negative reviews and citations from review sites. They're also the result of concerted promotions (but by linking pages, not by AdWords ads) that include the domain name.
For your needs, citations do have to include your domain name. Obviously, with regard subdirectory suggestions, the urls need to exist both on other pages and on your own site as well. Again, this isn't very common. When you see a subdirectory suggestion that's in relative isolation on a site, check it out and you'll see the combination of links, titles, and quantity of referring pages that brings such a suggestion about.
| 4:25 pm on Oct 9, 2009 (gmt 0)|
The italian Nicola Briani did a test in the past months to remove some results from Google Suggests that afflicted a customer.
It was a problem related to keywords published in many article marketing and press releases. So finally yes, it is possible manipulate Google Suggest :-)