Forum Moderators: open
However, it is quite obvious that at Webmasterworld, typical posters tend to be commercial webmasters, and SEO types. The very home page of this site proclaims "News and discussion for the independent web professional". Which would suggest that most y'all just ain't the people Google are trying to please. Google is aimed at the average Joe Searcher out there, who probably is just looking for information on a topic of interest to him. Maybe someone trying to settle and argument, find out more about his personal hobby, or some kid doing a paper for school. I almost never use the Internet to find things to buy, and probably this rarely is what most people out there searching have in mind to.
This leads me to the hypothesis that in the latest Google index they have intentionally altered the algo in preference to what this average Joe Searcher wants from Google. Yeah, this has mucked things up a bit for commercial webmasters and their SEOs, but you guys are just collateral damage for the greater good. Besides, if you are selling something on the Internet, that is what Google Adwords are for. ;) The more relevant the SERPs are on commercial searches, the less the average Joe Searcher will need to click on a Google Adword. Thus Google has no particular incentive to want the algo to be most relevant for commercial searches.
Now I ask: how many of you are really having problems finding the content you want when doing searches where the relevant results wouldn't be businesses hawking widgets and stuff? From what I believe to be that changes in the latest algo, they don't seem to be of a nature that they would tend to make it less relevant on non-commercial searches. Am I wrong about this for non-commercial Google searches?
I'm not sure I understand this sentence since it seems to contradict what others have complained.
I do understand that people in specialized areas may be being overwelmed by general/large/dmoz-listed .edu and .gov sites. If those large sites are not really relevant for the search query that is a problem. But if Yahoo put my keyword on its main page, how would you expect a search engine to *always* figure out that pagerank10 site with jillions of links isn't relevant if it says it essentially says it is. I'm not saying that a tweak isn't called for, but it is a complicated tweak where *relevant* themeing (not just site linking) is more highly valued.
Google reacted to what any objective person would have to say were a couple of serious problems that degraded the search results this summer, and in the case of selling pagerank, looked like they would have hurt hugely in the immediate future.
In my opinion, in most general search areas, especially the most competitive ones where the most abuse was taking place, the changes Google made this month were outstandingly positive moves that were absolutely necessary. Regional and specialty sites seem to have borne the brunt of these necessary changes, which means Google should look at that effect and tweak.
But, it can't be stated strongly enough that the selling of pagerank was insidiously non-helpful to user-searchers. Fixing that problem seems to have caused other problems in some categories, which now also need fixing, but if people don't see the large positive for search-users as a whole from the recent change I don't know what to say, except don't just look at your own fiefdoms.
So, this was as 'pure' an example as I could find without any overt manipulation of the results that I could see. Hopefully it will help with their analysis.
The best that I can say is that one of our "official" "dot gov" sites is still #1 as it should be. But in the past there used to be another page from the "dot gov" site in second position, (which was a COMPLETE mystery- why THAT page?) (It was indented, whatever that means.) We never knew, but now it's gone. Our campaign sites are now in second which is proper. One site has a flash intro page and now the first text page (the real "home" page) of the site is now listed sixth. (See what I mean about trying to explain? It's like asking me to talk without moving my hands. )
In a way. we get a bum deal from Google because our national party has a campaign site with VERY heavy traffic, but it uses dynamic pages which don't show up on Google backlinks. I suspect, if that link “counted”, we would rate higher in some searchs. C'est la vie.
It's REALLY weird in some other areas (Here's where it gets hard to explain...) If you click on "similar pages" the Flash file for one site is listed all by itself (Not with the page that contains it) Second, really old content shows up and so on. "Similiar results within 'our site' " also gets wierd, very wierd.
As long as Google uses a secret formula AND changes it without notice periodically, so called "SEO' is just voodoo. It's "Magic Numbers" straight out of Dilbert. (Faulty assumptions times incorrect data = Magic Numbers)
I'm just gonna keep improving my design skills, my digital photography skills and my web techniques. I create the best pages that I can and promote them to people who matter and leave all of the hand wringing to the spammers, and various other cyber grifters who actually BELIEVE this junk matters in the Big Picture. ( I guess it does if you’re selling widget enlargement widgets and time share widgets on widget island.) I care about what I do, and it Matters. I want as many constituents as possible to hear our message, but this forum gets pretty neurotic. I mean, I have a step son on a destroyer in the Widget Gulf. I care more about SCUDS than I do about SERPS, for gosh sakes.
Oh, and lately, despite our steady rise in traffic and improved Google ranking, other "referers" are surpassing Google.
[edited by: politicsandlabor at 4:34 am (utc) on Oct. 6, 2002]
[edited by: WebGuerrilla at 6:06 pm (utc) on Oct. 6, 2002]
[edit reason] TOS #25 [webmasterworld.com] [/edit]
I think that there are big problems with your results. Let's say that there is a music venue in Elboniaville, Ohio called "Little Siblings." And let's say you enter "little siblings elboniaville ohio" as search terms. One would expect the Little Siblings web site at "www,little siblings.com" to be number one, right?
Wrongo, bucko.
Three or four commercial gateway (city directory) sites rank higher. How could ANY site, no matter how well optimized- rank higher than THE "littlesiblings dot com" web site (and it's the real one, not a spoof or imposter)?
But I notice that it happens fairly often.
[edited by: politicsandlabor at 4:29 am (utc) on Oct. 6, 2002]
Easy. What if the venue "Little Siblings" didn't optimize well for "little siblings elboniaville ohio"? Like maybe they only mentioned "elboniaville ohio" once on the home page. Particularly so if "littlesiblings dot com" happens to have a low page rank, and those three or four commercial gateway (city directory) sites happen to have high page rank. The Google algo would be interpreting the "little siblings" part of the search as the unimportant part, and the "elboniaville ohio" as the important part, because those commercial gateway sites have high PageRank. Google has always been that when the search matches multiple pages, the high PR one beats the lower PR one.
Also, so long as that venue comes up in the top 10, the average surfer will know which of the 10 to click. Now, if that venue didn't come up within the top 100, then I'd wonder about the algo.
http://www.google.com/contact/spamreport.html Mention you've been reading on webmasterworld. Mods, I apologize for posting this link so often, but I keep waiting for people to take me up on this offer and only a very few have.
GooglyGuy, i apologize for posting the same spam so often, but I keep waiting for google to stand by their offers!
I reported the same blatant spam the 10th time now - maybe you don't understand german search phrases!? ;)
thanks Google.
zeus
P.s But who knows what happen next month?
I re-did some of the in-depth searches that I've done in the past 6 months. (for: an obscure variety of chinese green tea, a certain firearm attachment, modifications available for my specific car, salt water aquarium info, pumpkin carving patterns, several music genres, mandolin types & tuning, etc)
In these types of searches, the current SERPS are now clearly superior to the past.
I also searched deeply in my own website's category ( health ), and all the familiar sites are still there, just re-ordered somewhat. But the big improvement was the disappearance of multi keyword1-keyword2-keyword3 domains that were redirecting to on-line pharmacies selling viagra, growth hormone, etc. That's an improvement too.
I also searched "location hotels" for the past 4 trips I've taken, and I had no trouble finding sites listing hotels, prices & details, in many of the cities and towns I stayed in.
In summary, I personally think non-commercial searches are a little better.
But what are the best results for "mark"?
My impression also. I basically just do non-commercial searches. It may be that Google tweaked things to make non-commercial searches better (which are far more common), and this had the effect of making commercial searches worse. Also, it occurs to me with commercial searches, exactly what is "spam"? In a sense, all commercial sites are "spammers", and these are the ones clawing to get to the top with SEO. Precisely which merchant is the most relevant? From the point of view of the sellers, they'll all tend to think they are the most relevant. From the point of view of buyers, which is the most relevant? The cheapest? The most reliable merchant? The biggest merchant, in preference to the small businessman? If I enter "widget sales", and a bunch of sites come up that sells widgets, is that irrelevant results?
Looking at several such searches, this is indeed the case. "elboniaville" or whatever city will tend show up in link text for the elboniaville chamber of commerce, the elboniaville rotary club, etc. It is truly amazing how many sites don't have their target text on the page... and when they do, they sometimes have it in Flash or in a jpeg or gif.
Ironically, this is particularly true for marketing and image conscious companies, where you'd think that being located would be important.
I keep seeing this as a problem area for Google... it's been this way for a while... because those city names show up frquently in link text to irrelevant sites.
Sometimes, as SEOs, I think in talking about how relevant an algo is, we are talking about how well it suits our own optimizing styles, not just because we're being competitive, but because that's what we think relevance is.
now I don;t expect it to automatically rank as the top site for anyone searching for "suicide", "suicide helpine", "suicide prevention" etc...actually judging by the logs desperate people type some very odd things into search engines...but before the latest update a single word search for "suicide" (the most common as far as I can tell) would give a top ten entirely composed of major suicide prevention organisation (almost entirely the US based ones though) and a couple of good general info sites...we came 11th
now we have dropped to 23rd and several of the large organisations have dropped with us...and in our place are film, music and porn sites that just happen to have suicide in their title and domain name (and therefore in many link texts I assume)
I don't think this is particularly good...I understand why the algo needed tweaking in this way...and I have to get many of the links to us changed anyway since we are rebranding...so it's a temporary situation...but the results ARE poorer after this update
>Ironically, this is particularly true for marketing and image conscious companies, where you'd think that being located would be important.
Yep. It wouldn't surprise me that this "Little Siblings" music venue would have a Flash intro on the home page, or the name of the venue nowhere on the page in some nice, pretty .jpg created by some web designer clueless about search engines. It turns out that one of the few related sites to the main site I have in my profile was SO clulessly designed by some whiz kid who knows how to make pretty visually oriented sites ain't even in the top 100 for the main keyword it should be optimized for, and something like #70 for the other main keyword. And, these are very *uncompetitive* keywords! Yes, I am not kidding. The index.htm page is nothing but a .jpg, and of course not even an alt tag on that. No keywords in the page title, and heaven forbid that a single meta tag be allowes to defile the page.
This person who wondered why this elboniaville music venue is doing so poorly should probably examine that site's HTML code. Good chance the problem isn't with Google, but instead that the hired a web designer without a clue.
Why should this surprise you? I was most surprised that the main Usenet suicide newsgroup website (which basically is a how to guide on how to commit suicide) wasn't in the top 10, and was something like #50. I checked that site's home page. While it deserves to be in the top 10 for "suicide" based on relevance, it is really badly SE optimized. Why would you think for a search on "suicide" based on relevance that the top ten be entirely composed of major suicide prevention organizations is how it ought to be? I just checked the Google SERP for "suicide". 8 out of the first 10 listed were suicide prevention sites. I'd hardly say they are underrepresented. You're looking at this from a value judgement perspective. Based on my values I was disappointed that the main how to website on the Net wasn't in the top 10, or a website about Jack Kevorkian. I actually once shook Jack Kevorkian's hand. However, I realize the reason why the Google SERPs are as they are has nothing to with the algo, but is just a matter how well optimized the various sites were.
I'd have no problem if there were pro-euthanasia sites there...and at times the alt.suicide.holiday FAQ pages have been ranked highly...my problem is that sites where suicide has been used as part of a brand name outdo sites where it is actually the subject of the site
a good search is one that delivers relevant content...and relevant content to a single word search on "suicide", by most definitions, does not include unclad goth girls or a Belgian band
I just did some checking. That unclad goth girls site has a PR of 6, and is showing 347 backlinks on Google. My guess is the fact "suicide" is in the domain and site name means that people who are linking to that site tend to do so with "suicide" in the anchor text. That Belgian bands site has a PR of 5, 45 backlinks showing, and the band does have "suicide" in its name, so probably that is appearing in the anchor text linking to it. BTW, are you aware on a search for "stones", the top 5 listed sites are about the Rolling Stones, and not broken piece of rock?
Thus, on checking it doesn't look like those 2 sites you mentioned seem out of place on a search for suicide. Not all sites relevant to the word "suicide" are just dealing with the topic of killing yourself. These sites aren't some kids little home page that nobody cares about. They have good PR, and a lot of inbound links. They just seem out of place at first. One a close look, they do make sense for the query.
I must admit: the results look very good to me.
I don't have an answer for how the algorithm should be tweaked...I'm just saying that in this case the new version throws up anomolous results that the previous algrithm didn't
I'm not going to beat up on a search engine that has placed us top on a number of relevant two word terms even before I started looking at SEO...when a number of others are so hung up on an erroneous Look Smart directory entry that they think it's a site relating to New York City (we've had the entry fixed and one day MSN might get round to updating their database, I'm not holding my breath)
None of the above.
Is it really "optimized" for Google?
No. But it's not bad either. Does well on other engines that rely on "meta tags."
Besides, as happens so OFTEN on this forum, were missing the "forest for the trees." A search engine (any search engine) is supposed to SERVE the needs of web surfers, not the other way around. The mechanism that explains how Google could miss "Littlebrothers.com" in favor of other sites demonstrates a problem with GOOGLE!
First off, the Google "algo" should be massively weighted so that if you enter "search term" and "searchterm.com" is out there, that's Number ONE on the SERP.
Second, if I enter "search term city state" there should be a hierarchal weight to the first term (First term entered is heaviest, then the second then the third then the last is lightest)
AND some sort of special treatment for location terms as modifers of the preceeding terms. If I enter "search term elboniaville ohio" I am NOT seeking info about elboniaville, I am seeking to narrow my search so that I get the RELAVENT hit for "search term."
I've said it before, this is straight out of Dilbert. Google has secret criteria that it changes unilaterlally without public pronouncment. It's terms of service prohibit doing any large scale statistical attempts to "reverse engineer" those criteria. As a result, I have never seen any statiscally reliable emperical data on what effects changes in Google criteria really produces...
So, we have faulty assumptions times unreliable data equals "magic" numbers, which SEO types then sell (for cryin' out loud) to unsuspecting web site owners. Banum was right.
However, does it necessarily seem logical that people searching about "stones" would be searching for the Rolling Stones, as opposed to being rock collectors of sedimentary stones? If you check through Google, this is typical on one word searches of common words in the language. You get a mixed bag. In such cases, the user has to start using their brains. Like refine the search to "suicide prevention" or "suicide methods", etc.
Thus, the flaw politicsandlabor is not in the Google algo, but in your spelling. ;)
I read you opening thread with interest and I agree that Google is under no obligation to please SEO'S. Your hypothesis (above) is interesting and probably fairly accurate.
There is, however, a much more sinister explanation for this change.
I read an article in the (British) press today which stated that Google adwords are all but pointless because the SERP is so good no-one bothers clicking the adword links ... and no 'adword' click means no cash for Google.
Is it possible that Google could be doing what you have suggested purely to generate income? Playing with the algo so that commercial searches are generally less targetted and sometimes way off the mark so Joe Searcher clicks on an 'adwords' link and hey presto Google gets its cash?
I think Google is the best search engine by far. I also think they are clever enough to look after number 1.
Someone today posted that they got 1000 adword click-through's last month. They also said that they have beaten that already this month with well over 1000 hits. This would validate my theory.
You could pick a few choice 'commercial' keywords and reverse engineer the results to be slightly less relevant.
If this were true, would you still argue that it is in the best interests of Joe Searcher and for the greater good.
Food for thought. ;)
"I'm talking mass generated pro spam, designed to hit and run. That's what I don't see Google, or any other engine tackling...It's just that for my searches I don't see the advantage Google had in this field anymore."
In case like this, perhaps these hit and run pro spammers are popping up so fast that Google can't whack them down before they pop up more spam? Perhaps even Google is being overwhelmed by this now like the other search engines? It may not be that Google changed the algo and broke it. It might just be that the algo is now under attack by spammers to such a degree by the spammers that even Google now can't handle it. The fact that other search engines are also not handling this does go against the idea that Google had sinister or self-motivated interests here.
Google also has limited resources and money. I'm not sure, as someone who rarely does commercial searches, I'd like to see Google throw ever increasing amounts of money on dealing with commercial spammers. Doing so would take away resources I'd rather see devoted to improving the quality of non-commercial SERPs. Remember, neither the non-commercial or commercial webmasters pay any money to Google for them to generate relevant SERPs. Why should the non-paying, commercial webmasters (who are the minority of sites on the Net) get the vast bulk of Google money spent improving their SERPs? Google has no incentive to improve the quality of commercial SERPs until it gets to the point they are so bad Joe Searcher chooses another search engine. And, according to heini, none of them are better so this shouldn't be a great worry for Google that Joe Searcher will go elsewhere.
Now rfgdxm1, you wouldn't take things out of context and presenting them as authoritative at the same time, right? ;)
What people in this forum and elsewhere are complaining about is not that sort of spam I was talking about, so lets forget that.
It's all much simpler than that:
Google has "updated their algo", to quote GG, and many people think the revised algo is not as good as the old one.
It's on Google's side.
Everybody is free to judge on the new algo.
But when trying to evaluate and understand the new algo constantly talking about spam and spam reporting is offtopic.
? I thought they were? Note the title of this thread I started "What do you think of the quality of Google *non-commercial* searches?". Basically, everyone thought they were OK. So this update has only effected materially the commercial SERPs. Precisely what sort of spammers are people seeing other than the kind you mentioned?
>But when trying to evaluate and understand the new algo constantly talking about spam and spam reporting is offtopic.
No. Not if the complaint is "the new algo is letting in a lot more spam then it used to, and because of this it sucks." Specific cases of spam, or spam reporting, are off topic. However, the overall "spamminess" of the new Google algo, and trying to understand why that is, definitely is on topic.
I can see how spam fighting and reduced relevance could go together. If someone comes up with a way to put the most relevant pages at the top, and then delivers more than 50% of the non-paid search results, then people are going to mimic relevance.
For some of us, that might mean getting a small business ranked higher than its big-brand competition. For others, it might mean flooding Google with thousands of highly targeted, but useless pages.
Therefore, one way to fight spam is to put a little less emphasis on the relevance criteria.
I'm not saying that this did happen, just that it could happen.
I don't see more spam for the non-commercial searches that I can compare in this and the last index, but I do maybe see less relevance. For the commercial searches, I'd say moreso.
You've got a good point about my spelling. I should have tried it both ways. ( I honestly don't know how Google handles apostrophes.) Interesting, veddy interesting!
Try this before the mods change it: Try Tim O'Brien and then Tim Obrien. (There's a novelist, a musician and an insurance agent) notice how the results move up and down.
Veddy interesting...
But I stand by my rant that "city state" at the end of a query string should somehow be treated as a modifier. (Although I admit that I know that I'm not explaining myself very well...)
I do a website for a "Habitat for Widgets" chapter and this issue comes up there as well. Right now we've got LOUSY Google results and I'm not making much headway improving them, in no small measure because our biggest "friendly link" is the International site which is asp (or php) generated (I can't remember) and doesn't show up in our backlinks.... IF that still counts any more.
