Forum Moderators: open
That PR hogging theory is interesting
Googles logic may well be that this page must have some authority in the topic of the search phrase based on the fact that you are less likely to lie in the link text. You may wish to chuck garbage at the bots - you are less likely to chuck nonsense at your users!
mc still shows the serps with the "-" between keywords. I think that mc is not a fully active datacenter yet
>> links
Finally someone else noticed... wrote that somewhere in part two of the Florida update thread.
claus, didnt your post on local rank take into account outgoing links to similar sites in the same serps? This would bring directories back into play while knocking down the smaller niche sites that wont/dont link to competitors.
Term Links that have fallen out of the top100
Exercize 0
Buy Porn 90
Prayer 0
Vaigra 63
Personality 4
p***s enlargement 90
Education 2
Playstation Cheats60
Golf practice 1
Golf clubs 83
Anyone still doubt that there is a hit out on commercial sites?
Term / Links that have fallen out of the top100
Exercize 0
Buy Porn 90
Prayer 0
Vaigra 63
Personality 4
p***s enlargement 90
Education 2
Playstation Cheats60
Golf practice 1
Golf clubs 83
Anyone still doubt that there is a hit out on commercial sites?
Me. Small wonder that even the tiniest of algorithm adjustments could make 70 sites fall out of the top 1000 for "buy porn." There are probably 50,000 sites out there trying to sell porn, all competing for that top 2%. Any little change to Google's ranking system and there are thousands of new sites leapfrogging one that matches Google's new criteria a little less well. It's the law of the jungle.
I think the question is not "why does rfgdxm1 dislike commercial sites" as "why do you dislike informational ones?" Seems like any time a search returns a result that's not about buying the widget in question, you guys think it's not relevant. Well, if you search for "widgets," widget sales are relevant, but so are photos of widgets, how to build widgets, the history of widgets, and the Widget Appreciation Club. Why should ONLY e-commerce sites come up? Not everyone's looking to buy something; in fact, most searches are probably informational in nature. *shrug*
rfgdxm1, not this time. Without actually writing out the product name/keyword phrase it is a little difficult to show you, but I assure you that there is no need for anyone to search for this item that did not want to obtain this item. It does not lend itself to discussion on its particular merits - it is very straightforward. When Joe Surfer looks for my industry's 'professional doodads', he does so because he needs to get his hands on one right now.
There are no 'professional doodad hobby groups', no 'history of professional doodads', nothing to repair, no safety issues, no forums reviewing the particular merits, etc. People do not browse idly for the doodads I sell - they are looking to buy a necessary evil for their profession. They don't buy my doodads because they want to, they buy them because they have to. They want to get it, use it, and cross the damned thing off their list so they can actually do what they are good at.
And when they put in the phrase 'professional doodads' in Google, right now they cannot find a single purveyor of such within a reasonable number of page results.
That's why agreed with the poster that pointed out that webmasters understand the particular search niche their industry occupies very, very well. We need to know, more than giants like Amazon, *exactly* what our market segment is looking for, because we cannot afford to miss the target.
You don't know everything about everything, dude. Sometimes other people know stuff you don't.
And this is a lot of money we're talking about here, don't know the current price expected, but I read upwards of 1 Billion dollars, so if you can nudge that number up by say a few percent, well... why would you not do that if you were in google's position?
Personally, I would assume that everything google does until the IPO happens is done only to accomplish this increase in profitability, that's what any halfway competent chief executive officer should do, and it's pretty clear that that is what they are doing. That's what they get paid for.
the real question is why would google not do this at this point in time
Because it would be misleading. Omitting commercial sites in favour of AdWords would be as bad as including paid listings without labeling them as such.
Not so sure about that. 3 of my sites which have been dropped for main keywords all are info sites with plenty of outbound, unpaid links. Surely not on every single page, but probably a ratio of at least 2-3 outbound links per page
I just got back from an SEO bootcamp, of course this was a topic we spoke quite a bit about.
Our discussion brought up the fact that Google seems to be doing quite a bit of spidering, almost every day, and they are most likely doing an ipo very soon.
1. Wouldn't it make sense for Google to have their own directory, not pulling information from other sites?
2. With millions of websites this would take some time.
3. This "Florida" update is creating quite a "buzz" inside and outside the industry.
4. If they completed a "new" directory and all the "new" results were posted, and they were the best results from a completely proprietary source, Google would benefit the most from this.
pretty good marketing for an ipo?
What are your thoughts?
In my industry, the generic product name is the specific keyword phrase. The results, although they contain the keywords in question, are utterly useless to anyone using the search term to actually look for the product that *is the keyword phrase*. Joe Surfer looking specifically for 'professional doodads' has a reasonable expectation that he will find sites selling professional doodads, not pages that have the words 'professional' and 'doodads' somewehere in them but have nothing to do with professional doodads except in the most ephemeral manner.
Wrong. The person may have no ineterest in buying. Perhaps he is looking for information on the safety of professional doodads, or how to repair them, etc.
rfgdxm1, "there you go again" making another blanket statement about an industry which you have not even been made aware of yet (since none of us know what industry is being referred to in that post).
Wow, talk about being a member of the google apologist society;) Just kidding, lighten up people! ;)
Seriously speaking though, you are defending google and apologizing for google even in instances when you have not been told of the specific industry, and I think that is not logical to attempt that task.
The original poster who mentioned that "professional doodads" are sought by people needing to "buy them" could very well be correct.
If "doodads" are the focus of that person's own industry then I'd be willing to accept that person's real world data over your "blind speculation", since you are not even sure what "professoinal doodads" refer to so it would be ludicrous to accept your blind speculation as "the truth".
You are making 'blanket statements' in instances where you are not even sure of the "specifics" of a given argument or topic, you are trying to counter an argument without even knowing any of the specific facts.
Guess what? It may be true that "professional doodads" are sought by people mainly looking to buy them 99% of the time, that MAY be true for that specific keyphrase (and the fact that you would never acknowledge this possibility is what makes your point of view biased and slanted toward google regardless of what the facts are).
For example, if I wanted "cheap doodads" (which I don't) and I typed that into a search engine, I would expect to find places to buy cheap doodads, I would not want the "history of cheap doodads" or a document discussing how doodads are causing global warming, etc.
Why should people listen to my data above?
How can Google know for sure what somebody wants in that keyphrase above?
Because the word "cheap" gives the context of the search in that instance, and that's not rocket science either, just basic logic.
In fact, the most common reason for typing in a term such as "cheap doodads" is to BUY THEM, there are other obscure reasons to type in that term but the MOST OBVIOUS reason is to BUY THEM. Anybody who disputes that simple fact is really not evaluating the situation with a clear and realistic mindset.
The word "cheap" gives OBVIOUS "context" to the search, and many other industries and products have other (different) words which associate OBVIOUS context with a particular search (just because google is not intelligent enough to recognize such "context words" doesn't mean the context is not present in the searches).
Well, not unless you are a google apologist looking to side with google before even reading all of the facts;)
Obviously not all keywords are being sought by "buyers" (nobody is saying that so please don't repeat that issue again, it's been repeated enough already as a "red herring" to this debate so let's deal with the facts;)).
Many keywords do have an equal (or greater) number of people searching for "general information" and not information about "purchasing" an item, but there are PLENTY of terms where the OBVIOUS reason is to mostly BUY what they are searching for.
If you want to pretend that in all instances there are an equal number of searchers not looking to buy something then go ahead, but that's just not a realistic mindset because that is not true in all cases, and it's not necessarily true even in most cases, but could be true much of the time (It's speculation to try and debate how much of the time this is true, maybe more than 50% or maybe less, who knows).
But it's about recognizing the context of a search and recognizing that certain words typed in by the user give away what their CONTEXT of the search really is, it doesn't take a mindreader to see this either;)
Finally someone else noticed... wrote that somewhere in part two of the Florida update thread.Yup, when I added my altavista links I also added a link to my biggest competitor. Besides, its really not a big deal as my home page is now less valuable than my links pages.:(
Sites that are similarly optimized (at least to google's eyes) are appearing to being filtered out, except when adding the -asdffdsas or such filter, which returns the "normal" results. I say there is nothing one can do since your competitors are what cause you and themselves to serp out. Unless you majorly change your seo strategy such that it does not look like your competitor sites, then all similarly serping sites with similar seo pratices will be filtered out.
Sites that are similarly optimized (at least to google's eyes) are appearing to being filtered out, except when adding the -asdffdsas or such filter, which returns the "normal" results.
That is one of the conclusions that I have been coming to. I have done a lot of analysis on one particular search phrase where all the pages in the previous top 50 which repeated the exact search phrase more than twice (no correlation to H tags) have been removed - except for sites that use the exact search phrase in anchor text in outbound links. However, for other search phrases the criteria for the "filter" are quite different. I think Google is looking for patterns in the results and filtering similarly optimised pages.
Heads up for changes at the dc's. Rumors swirling that there have been some changes today momentarily
Just a quick check on thread topic. Four-worded highly commercial phrase w/ adwords top and right; "cheap web site widgeting" - total no. of results for standard search:
www: 2,380,000
www2: 2,470,000
www3: 2,460,000
-ab: 2,270,000
-cw: 2,160,000
-dc: 2,450,000
-ex: 2,040,000
-fi: 2,150,000
-gv: 1,570,000
-in: 2,270,000
-kr: 1,480,000
-lm: 1,450,000
-mc: 1,580,000
-sj: n/a
-va: 2,380,000
-zu: n/a...about a million in difference top-bottom. GV returns differences in top 5 - MC return differences at #6 and below. LM and VA has switched #7 and #8 relative to the other DC's. KR has switched #8 and #9.
Terms "web site" interpreted as one word - Top 3 matches "cheap web widgeting" in title (which is right for this industry, in fact used more often). Except for GV, #4 is "cheap web site widgeting" and #5 only matches "widgeting" in title.
All datacenter results are relevant (the whole Top 10 can lead you to a purchase) 6 may be considered "directories", 2 of these are "review" sites, the four remaining are sellers of this product. All sites are SEO'ed, using everything in the book (in this industry, that's quite normal).
Added: if this is significant, i'd personally call it Florida round 2 and not Galen, it's too early for an entirely new one.
/claus
[edited by: claus at 2:49 pm (utc) on Dec. 3, 2003]
i see a site that is basically about 40-50 subdomains divided into regions and then linked as "discount region widgets"...now he has swapped links by getting sites to place a link to each subdomain from a single page...every links says discount region widgets so now the large number of links makes the link page keyword heavy for "discount"...so all in bound links to all subdomains which all cross link are weighted as "discount" relevant...he has kept top slots across all subs with no content and cross linking for all discount related searches....i think this givs a big clue to how the anchor text filter is working....have a think about it...!
Soapystar, first, don't you think that it's possible that all of the above is simply related to the fact that subdomains had huge success in Florida?
Also, I'm sure that having inbound links for each subdomain is good. Then of course there's the cross linking, which didn't seem to hurt most sites except for the homepage.
Do you believe there is more going here, e.g., a more subtle point than I was able to take from your note?
Also, how's the homepage doing for this site?
TIA,
caveman
I personally am leaning to the notion that G is building a new index and using an old base for now, testing as they go. That helps explain why some older sites seem to be having more success ... generally speaking. IF that's so, then it's hard to know where we'll be in a few weeks. Hopefully we'll be seeing some MIA's back soon. Just hope the mom and pops can hang on.
There is little question in my mind however, that the spammier categories will never return to pre-Florida status, and that's fine, since there was a lot of junk out there. Unfortunately G threw out a huge number of quality sites at the same time...including quality affiliate sites (guys, even MSN has affilate links).
world3d, yep, lots of those kinds of comparisons to be found. Key questions remain, however, like what was the motivation, and what was the method? IF the motivation was spam attack, and not an attack on commercial sites per se, and if some sort of bayesian filtering is a part of G's 'solution', then it's only natural that more commercial categories would see more casualties.
But as many have posted, in addition to seeing some very good commercial sites get blasted, thousands of non-commercial sites, were hit also. That's important to remember, I think, because it seems to rule out the notion that a simple list of 'bad categories' is what's going on here.
I do know understand way my swedish indexpage is lost,
though company name is in english and is irrevelant to a swedish site............... and the spanish is not lost but it lost position in first 100 but searching for the english keywords, not for the spanish, these are the name of company (city+rentals) these words are irrevelant to a spanish site..................
[edited by: helenp at 3:43 pm (utc) on Dec. 3, 2003]
MQ
I agree with your theory - looks like it from what I am watching, explains why the search words in my logfiles seem to change daily.
>> claus, didnt your post on local rank take into account outgoing links to similar sites in the same serps?
Links yes, outgoing no. It was "incoming links from" - still, these can't be there unless other sites in these SERPS have them as outgoing. A "reversed localrank" perhaps? Definitely an interesting theory...might explain directories ranking high, but will not explain non-outlinking sites ranking high. Real localrank (not reversed) might also be good for directories, however, as other topical sites tend to link to these...
/claus
