| 9:28 pm on Sep 29, 2002 (gmt 0)|
Any way we can keep GoogleGuy from seeing the results? ;)
Otherwise 2 months down the road we'll be having another poll!
Seriously, I don't see much mileage in a simple poll. I couldn't answer question 3 if I wanted to. The answer is "a bit of this and a bit of that".
I think we just need to see as many theories as possible and test them on our own sites. The ones that most people agree with will probably be close to the truth.
| 9:51 pm on Sep 29, 2002 (gmt 0)|
not that they're doing this...but is it possible that each industry/category has it's own set of rules/algo?
| 11:47 pm on Sep 29, 2002 (gmt 0)|
I'm going to start with this response. Of course since I put this together there has been more discussion and I'm going to put a comment on that together because I believe it might be useful.
|All my internal links to the home page contained only the word "home", whereas the site which stayed top has links to "Location1, widgets, location 2, location 3". We just did this last month to make sure we maximized the value of our internal links. - SlyOldDog |
Bravo you SlyOldDog. I remember discussions on this not to far back. We just updated navigational links as well and I believe that contributed to this site doing so well this round. Finding that mix that works for both the visitor and the engines is worth the effort perfecting.
|This suggests anchor text on internal links is key - SlyOldDog |
Whatís interesting when looking at results like this is how what seems really important one round becomes less important in another. Thus, my belief in creating well-rounded sites, balanced in all the key areas and always pushing that envelop a bit. With internal linking it really surprises me how little time and effort people put into working it. Changing link text is a great method for testing.
|Our sites which dropped have FAR less incoming links, again suggesting that the amount of incoming external links with good anchor text is key. - SlyOldDog |
Iím not sure I follow the logic here.
What some folks may not realize is that when you conduct steady research over time you begin to see changes that others see as happening suddenly have actually been working over time. When things hit that critical mass changes often appear sudden but in reality have experienced subtle shifts for months. People see it as sudden and I see it more as trends that have been developing.
I too have seen what you suggest in #3 SlyOldDog.
|the most frequently updated and freshest results - Visit Thailand |
I agree. This too has been building over time in importance and those sites that more obviously stay fresh and update frequently are finally pushing forward. I try to revisit pages on my site frequently and when I can make a good change site wide I do it, especially with the links. Just keeping both internal and external linking error free and all images, that monthly maintenance we should all be doing, helps keep a site fresh.
|It would explain why one site of mine that I havenít touched in over 3 months goes from strength to strength every month, and others, for which I have actively been seeking links for have suffered slightly. - subway |
Thatís the flip side. I have seen this happen with older sites that were basically in good enough shape to let be for a bit. I found that on an average though Iíve had better results if I at least perform regular maintenance each month. Also, if youíre testing things, even a site that is doing well can be tweaked. I donít think folks spend much time with their logs and that surprises me. Thatís where I get the ongoing information I need to keep improving on my sites.
|Google has apparently discounted the importance of backward links (maybe only in competitive areas) and paid special attention to any backward links that may have been gained by unsavoury methods. - subway |
I just donít see that. I watch over a few sites I inherited and ugh, the backward links are full of stuff we warn folks about and they continue to do well. It may be the fresh linking I bring in pushing out the Ďunsavoryí lot.
What I noticed is the structure and arrangement of the backward links. Bravo Google because they are much easier to follow this round.
|dynamic pages have been treated more kindly this time around. - taxpod |
Actually taxpod Iíve noticed this trend building now over the past three months. Initially noticed through backward links and then building to this month and showing up more in the serps. This may have affected a few searches. I wonder if others have noticed this or could look around for their keywords and report back if they see many.
|My interior pages are getting loads more traffic, especially the dynamic ones - taxpod |
Thatís how it was nearly two years ago. Then it was because of the dynamically generated directories and portals with pages showing up all over the results. I wonder if folks are seeing that again.
hooloovoo22 Ė Welcome to Webmaster World. Thanks for contributing.
|I'm still leaning towards a randomizing 'feature' being toyed with. Ė ScottM |
Well if we all picked a keyword from our industry and set the search results to 100 and took a snapshot of that each day, then followed through with any changes, noting when those changes happened and what happened and then comparing the results as keyword/phrase 1, keyword/phrase 2, so on. I dream of that.
Anyway, Iíve taken all my snapshots but havenít had the time to analyze the results Iím following month to month. Just thought this was an interesting thread to participate in.
|Now, this works great for you this month, but lets say that next month- john316 |
That surprises me John. From what I see, following steadily along as I do less jumping back and forth and monthly changes then people seem to think. I see changes implemented and taking months to move through and then dominating an industry. These shifts are subtle until suddenly their noticed. Itís something different each month. Thatís not to say there havenít been major changes that wrecked havoc suddenly. What I see happening right now are tweaks that may have occurred months ago finally pulling results up in a way that significantly impacted enough people to hear them raise a notice. You can tell by points made here, trends Iíve been following for months are suddenly a Ďbig change in the algoí.
Research folks. When you notice something, if it grabs your attention then track it for a couple of months and report in. Someone else may be tracking the same thing or perhaps were suddenly hit by a big algo change. Thatís when we can put our heads together and see if it was sudden or building gradually. Following the track of research can seriously help you anticipate changes in advance of them taking place so you can move your eggs around.
Too many wordsÖTo little timeÖ Regards to allÖ..
| 11:57 pm on Sep 29, 2002 (gmt 0)|
With regards to the health results changing. Consider for a moment a particular industry, health in this case, several months ago having a SEO come in and organize and then get listing nearly a hundred university (edu) sites that had not been or were under listed previous to that. Now lets just imagine those sites working there way through ODP, through Yahoo, through Zeal and because of the content and the results slowly showing up in the results so that other sites begin to link to them. Imagine what that would do to an industries results over a period of time.
This is how the work we do as SEO and conscious webmasters, designers or whatever impacts over a period of time. Thatís a key to how I try to influence change. Itís still a keyword/phrase at a time but under the umbrella of the industry. In other words Iím finding the results I want by breaking it down to the keyword after Iíve figured out the industry. Take command of the industry results and then feed them the keywords backed up by relevant fresh content, but heyÖ what do I know?
| 3:25 am on Sep 30, 2002 (gmt 0)|
Ok, Iíd like to bring this discussion back so here are some things I am following as a trend.
Sites being cleaned out that were ranking for a keyword when the only related use of the keyword on the page was a link out. So, if you have been receiving traffic for a keyword that is only used once in the keyword text of the page and thatís heading out, well I think Google is filtering those out of the SERPS. This may not be the intention on Googleís part simply a side effect of some other tweak a few months ago. For whatever reason itís been a slow weeding process just this time a bit more obvious. For me I will continue to track this, looking then for that trend with each months results.
There are other things may still be keeping a page in the SERPS for a particular keyword and that includes the use of the same keyword/phrase used in the text leading to the page. Of course then if that page doesnít have the additional content to support that keyword then itís bound to drop out of the results, as it should.
Iím also seeing pages cleaned out where the only use of the keyword is on the css heading tag.
| 5:06 am on Sep 30, 2002 (gmt 0)|
Jane_Doe's impressions fit what I'm seeing.
So when she mentioned the fall of multi-keyword domains, I checked one network that was #1 or #2 for several of the keywords I follow.
The poor guys! I checked a couple of keywords and their sites weren't in the first 200. Total nuclear destruction.
I didn't save the results from last month so I can't check it, but are you seeing fewer multi-keyword-hyphenated domains in your top 20 results?
I don't know, however, if the multi-keyword domain was the problem for these guys. They had a network of maybe 10 domains all very small, heavily interlinked and with few - perhaps, no - independent incoming links. Perhaps the lack of independent incoming links caught up with them. The sites were created last winter and ranked super-well in Google ever since... until Friday.
It is, of course, only one network of domains but it does seem to support Jane_Doe's observation that multi-keyword domains are "mostly gone."
[edited by: Learning_Curve at 5:40 am (utc) on Sep. 30, 2002]
| 5:38 am on Sep 30, 2002 (gmt 0)|
Cause and effect.
Multi-keyword domain falling may not only be due to the fact they are multi-keyword, but because of what many people who buy multi-keyword domains do! I guess they are mainly people with not much real content interest in the keywords, but SEO specialists who buy them for SEO purposes only. Hence much spam. Iid say a multi-keyword domain by itself may not be the problem, its just the results of the motivation for people buying/developing them.
| 5:44 am on Sep 30, 2002 (gmt 0)|
|Sites being cleaned out that were ranking for a keyword when the only related use of the keyword on the page was a link out. So, if you have been receiving traffic for a keyword that is only used once in the keyword text of the page and thatís heading out, well I think Google is filtering those out of the SERPS. |
Paynt - great observation. If it is indeed a global change what a great way for Google to put some of the relevance power back to normal text rather than links all the time. It seems a very smart way of getting rid of much scam and spam. Who wants to go to a page that just leads you to another page anyway? I want to go directly to a page that has content about the subject, not one that directs me to another page that has content about the subject.
All power to non-linked text! I had a feeling that Google was maybe overdoing the emphasis on link text anyway. Taken to it's extreme, this could mean more hits to interior pages of sites (that's good for the searcher! - they can move from the specific to the general rather than the other way around) and lead better to actual original sources rather than second parties who quote them.
| 6:09 am on Sep 30, 2002 (gmt 0)|
I disagree strongly with people who say anchor text is less important in this update. Of course I base this on only one site which I have been monitoring for over six months.
This site has been dead for nearly tow years and now contains absolutely no relevant content. This site has moved from 10th to 4th solely on the power of not updated links containing the relevant keywords in the link text. The keywords are nowhere to be found on the site.
This seems to be in total contrast to the supposedly lower importance given to the anchor text.
| 6:16 am on Sep 30, 2002 (gmt 0)|
>Multi-keyword domain falling may not only be due to the fact they are multi-keyword, but because of what many people who buy multi-keyword domains do!
I agree. There's a difference between between association and cause and effect. There may be a noticable drop in multiple keyword domains (MKDs) in some fields, but that might be because multiple keyword domains are frequently bought by spammers and/or people who put up networks of smaller sites, and not because Google has penalized MKDs per se. I still see MKDs ranking well for some types of search terms.
| 6:19 am on Sep 30, 2002 (gmt 0)|
Bikeman, do the pages that use link text to your site also repeat the link text in their page without being linked? I guess this is a primitive form of theme based link authority.
| 6:21 am on Sep 30, 2002 (gmt 0)|
Folks, I can tell you for a fact that keyword-in-URL-domains I have were not impacted at all. But then I quit doing much interlinking after last December's PR0 disaster. These domains exist in highly competitive categories. Almost all are very well ranked. Other factors are in play here.
| 6:43 am on Sep 30, 2002 (gmt 0)|
The site I am talking about is not the one in my profile.
As far as I have seen the majority of sites which link to the site I am referring are link farms and do not contain the keyword elsewhere in their text.
I am trying desperately to understand why this site is ranking so high and the only reason I can find is keywords in the anchor text.
| 6:52 am on Sep 30, 2002 (gmt 0)|
Have you ever considered that what caused that site to go up, is that the sites above it went down?
| 7:08 am on Sep 30, 2002 (gmt 0)|
The sites which went down have relevant content and links from themed sites and higher PR pages.
Can't see why they would go down the rankings.
| 7:09 am on Sep 30, 2002 (gmt 0)|
|Folks, I can tell you for a fact that keyword-in-URL-domains I have were not impacted at all. - Beachboy |
That sugggests it may have been a devaluing of their internal links. Or perhaps more likely since they dropped out of sight, Google lowered the threshold on something which caused this particular network of keyword-in-URL-domains to have their internal links disallowed altogether.
| 7:21 am on Sep 30, 2002 (gmt 0)|
|Sites being cleaned out that were ranking for a keyword when the only related use of the keyword on the page was a link out. So, if you have been receiving traffic for a keyword that is only used once in the keyword text of the page and thatís heading out, well I think Google is filtering those out of the SERPS |
I have not found this to be the case. Imagine a timeshare resort that has one site to sell lodging, vacation rentals, etc. And timeshare is purposely not on the site ANYWHERE. The same company operates a separate real estate site to sell the timeshares (no duplicate content). This month I changed the link FROM the lodging site to include the keywords "timeshare resales location" and now the LODGING site ranks very well for that search term. I changed this after the last major crawl and it was found in the freshness crawls. It may be important to note that the text was alt text of an image link and it was added to every (six) page on the site. The sales website does not reflect any changes as it was after backward links were indexed. However, it may not be that the outgoing link text is being weighed the same. This site has a higher pr than the others and that may be playing a large part as its not a particularly competitive search. Or maybe Google uses semantic spacing to know they are at least related to the timeshare industry. But given that we have tried to keep off any hint of timeshare from this site I can't say that content is important.
MODs if you can't post this I'm sorry as I can't really come up with any other way to say it. They are all very general categories. If not please post (or read anyway) the rest.
GoogleGuy posted somewhere around page 10 of the Sept Update thread (or Google is horrible) about reporting searches that were coming up with less than satisfactory results and what to do. Perhaps if that were a separate thread with GG's name next to it it would be more effective.
| 7:54 am on Sep 30, 2002 (gmt 0)|
I wish I could comment on the possibility of devalued internal links. Most of my sites feature flat menus, and I go to a lot of effort to assure inbound links (from other sites) to various internal pages. As a result, most pages in most sites have links to them from other pages in the same site, as well as from outside sources with quality PR. There is close correlation between the phrases for which each page is optimized and the inbound anchor text from wherever. Consequently, even the internal pages have a strong presence in the SERPs.
| 8:44 am on Sep 30, 2002 (gmt 0)|
Not related to my previous post I find that when searching for "location vacations" the results are similar to as before the update until about #20. From then on its strictly big business. One airline affiliate after the other. There are three in a row with the EXACT same site but different URLs. I have no problem saying this is right before mysite.com and yes I think it sux.
| 11:20 am on Sep 30, 2002 (gmt 0)|
Best thing to do, is to think what you would do if you were Google.
The most clever boy of the class does not like to get teased that sometimes his calculations look stupid. [webmasterworld.com]
Hence all measures would be full-force against Googlebombing and Buying Pagerank.
I agree with NFFC that first of all - as much as they dislike it - they installed some manual penalties or filters such as the SearchK Pagerank sell thing and the recent "go to hell" bombing phenomenon.
The other measures seem to look something towards this:
(mainly from this thread [webmasterworld.com])
1. What I see with this update, is that if the anchortext text is not in some important part of the on-page text (title/H1 etc) of the page recieving the link, the ranking falls. ("Go to hell" does not exist on the linked to page). This is a logical step towards facing googlebombing.
The back-side of this measure would be that "good-natured-well-intended" Googlebombing would have less effect (e.g. linktexts to a page without text).
2. A one-off high Pageranked anchortext link equalling "search query" I had from a topical directory (not ODP) pushed me into first page two updates ago. With this update I'm on page three for that "search query".
One could say google gives you a temporary boost for a new quality link, but I would guess that the "logarithmic PR" boost of that mainly the anchortext in ranking has been tuned down. That is, an external incoming anchortext from a PR7 page containing "blue-widgets" towards a page with "blue-widgets" in all the important areas, is not worth the log factor to the power of three of the same anchortext coming from a PR4 page (not that I ever believed it was such a strong factor in ranking). This is a logical step against buying high Pageranked anchortext links.
In any case some more even distribution is put into place.
| 11:54 am on Sep 30, 2002 (gmt 0)|
My husband and I bought some books and learned how to "build a web site" and did so ourselves for our family business. I was totally computer illiterate to begin with and didn't find Webmaster World until after google placed us in www2/3.
We submitted our site to every search engine we could find. We guessed on what to put for keywords and title, and don't even have a description to this day.
Here is how we ranked with google:
*For our title, (when other people have the same title) we ranked #1 on page #1
*For keywords in our title, we ranked page #4 (this is for "hats" which I am assumming is a common SE keyword used)
*For the keyword that tells what our product is (this phrase is not in our title) we ranked page #2
So, from what I am reading in some of the other posts, I can't really complain. Especially for guessing...
| 1:56 pm on Sep 30, 2002 (gmt 0)|
|I didn't save the results from last month so I can't check it, but are you seeing fewer multi-keyword-hyphenated domains in your top 20 results? - Learning Curve |
Not in the results Iím looking at. Keyword dominated urlís come in a lot and now canonicals are working their way up. Sometimes I think a few folks out there must have been following ideas I have shared here. What I donít like to see though are sites still there and coming in where the keyword phrase cannot be found together on the page anywhere and are only there because of the anchor text leading in. I do see these sites weeded back out within two indexes. What Iíd like to know about these instances are if the site popping in for an index or two and not relevant to the results, if itís because itís a new site and new sites get a short term boost if the anchor text leading to it uses the keyword phrase. It then takes a few months to weed out from the results if as a website manager you donít notice the trend from your keyword traffic analysis and go in and fix the page to support the traffic.
Blah, blah, blah. I have four discussions Iím working on comments for, a couple of stickies and I havenít gotten through my email yet this morning. Sorry but Iíll have to read the rest of the comments to follow when I can get back here. Really want to respond to your question though Learning Curve because others have pointed out this observation as you have and I wanted to point out that my recent observations show multi-keyword-hyphenated domains in my top 20 results. If those are being weeded out then I might suppose it was a side effect of some other trend. I'll keep on eye on it though.
I think google is trying to fine-tune page supported relevant content. This has been talked about in the past and I see sites that have been working it moving up the ranks. In some industries that are less competitive than others this may appear a bigger change and therefore startling. Look around at posts after updates and there are always startled people. Those are the trends I start following with research. Anyway, moving onÖ..
| 3:16 pm on Sep 30, 2002 (gmt 0)|
|It is, of course, only one network of domains but it does seem to support Jane_Doe's observation that multi-keyword domains are "mostly gone." - Learning Curve |
See, I didnít read on and I should have, now I see I should finish this discussion before moving on. Iím sorry Learning Curve but I got lost in the first part of your post.
I think the key to what might be happening or at least should be looked at is the first part of your comment where you mention
|only one network of domains - Learning Curve |
Thatís what Iíd be looking at. See I think whole bodies are moving out of industries
Rarely do I find myself not understanding you chiyo but this may be one of them. First, I see where you are going with thisÖ.
|Multi-keyword domain falling may not only be due to the fact they are multi-keyword, - chiyo |
but then I donít understand thisÖ
|because of what many people who buy multi-keyword domains do - chiyo |
do what? Fall out?
|I guess they are mainly people with not much real content interest in the keywords - chiyo |
but I do get thisÖ Maybe not that they arenít ďinterestedĒ but certainly they didnít maximize on their efforts or fully follow up on their potential. People are not reading their logs nor do they know what to get out of them when they do and how very subtle changes can make a huge difference in results.
BikeMan Ė I want to make clear that I havenít had a chance to look at the results Iíve taken snapshots of for the backward links changes except to note that I do like the way Google is offering them up and arranging them this month. I will probably comment on it somewhere when I have time to analyze my notes.
I agree completely with Jane_Doe in message #44, so that explains where you were going Chiyo?
Then chiyo #45 to BikeMan, good question. If itís not happened yet I would be watching that. If a page is drawing traffic for some reason to a page that isnít supported by the content then either support the page with content for future updates or figure in a few months it will drop back out.
This is such a long discussion and Iím absorbed with it. Iíll post these comments and then finish reading again when I have a chance. This is really a good discussion, thanks to all for sharing so much information and detail.
| 3:54 pm on Sep 30, 2002 (gmt 0)|
These are the rest of my comments for this discussion...
Beachboy # 46 wish Iíd had more time in my last post to address everyone. I see you also havenít had problems with the keywords in the domain. Keywords in the domain isnít all that make a site spammy. It continues to come back to supporting a visit with content.
|The sites which went down have relevant content and links from themed sites and higher PR pages. |
Can't see why they would go down the rankings - Beachboy
Look deeper Beachboy. Take a snapshot now of the results you are seeing, the first 100 for the keyword and then do the same next month and the month after and track those results. Then come back and let us all know what you found out.
We all need to start backing up our theories and ideas.
|This month I changed the link FROM the lodging site to include the keywords "timeshare resales location" and now the LODGING site ranks very well for that search term. - Powdork |
Thatís exactly what I come to say eventually in a post this morning. Iíve noticed that very same thing and it holds for a month to three even before dropping out if the page isnít then supported for the keyword. Itís the follow through to watch now. Report back on this in two to three months without doing anything now to support that keyword phrase on the page. Or, tweak the page and keep it drawing the traffic. Some folks would rather not take a chance in the spirit of research;)
Nice comment Beachboy #52. Thatís a good strategy.
Vitaplease - #54, geez I wish I could take my normal Google update week off, there really are some great discussions going on and lots of information shared this time out.
Your #1 Ė I see that too.
Your #2 Ė nice post, Iíve copied to give more thought too.
| 4:30 pm on Sep 30, 2002 (gmt 0)|
The comment in the gray box in your post #58 that was attributed to me was actually made by BikeMan (post #49). ;)
| 5:11 pm on Sep 30, 2002 (gmt 0)|
but then I donít understand thisÖ
because of what many people who buy multi-keyword domains do - chiyo
do what? Fall out?
Im suggesting that many, but of course not all, of people who use multi-keyword-domains do so as part of a SEO venture. Having keywords in the domain has been seen as a quick way to increase rankings. So of late people may have been rushing off to buy these domains, no matter the subject and their expertise in it, and using them as just one technique in a bag of SEO tricks. I think a higher proportion of multi-keyword-domains as compared to older or non-keyword domains are being developed by people who are use SEO tricks rather than good content. Its a short term opportunistic strategy. Thats what i meant by "the things they do". So the culprit for any preception of the dropping of mult-keyword-domains in sERPS may not be the multi-keyword-domain itself, but the fact they sometimes come with many more "tricks" . I hope thats clearer. Maybe not :) And let me make it clear that im not saying all owners of these domains are shysters! Heck we came close to buying a couple a few years back.
| 5:14 pm on Sep 30, 2002 (gmt 0)|
Yep Jane said it much better than me!
| 5:20 pm on Sep 30, 2002 (gmt 0)|
Sorry Beachboy, thanks for pointing that out, geez but it's hard to keep up this morning.
| 6:01 pm on Sep 30, 2002 (gmt 0)|
re: multi-keyword domains
GoogleGuy recently dropped a hint that they are watching the numbers of domains owned by someone to try and tell the difference between SEOs and mom & pop shops.
Are the multi-keyword domains that have been getting dinged in that category that they now are watching?
| 7:12 pm on Sep 30, 2002 (gmt 0)|
|Report back on this in two to three months without doing anything now to support that keyword phrase on the page. Or, tweak the page and keep it drawing the traffic. Some folks would rather not take a chance in the spirit of research-Paynt |
Unfortunately I can't tweak it since we have to keep 'timeshare' off of it. Evidently, people find out they are renting from a timeshare resort and they think they are going to have to take a tour or something. That is not the case here. I will keep it that way and see if it drops off. I really can't wait for the next update to see how the new alt text helps the resale site. It currently ranks #1 when searching for the previous alt text. Of course, ranking #1 for 'click here' (not the actual alt text but you get the idea) doesn't get you much.
Is it too early to start the "when ist the next update?" thread? :)
| 7:55 pm on Sep 30, 2002 (gmt 0)|
|GoogleGuy recently dropped a hint that they are watching the numbers of domains owned by someone to try and tell the difference between SEOs and mom & pop shops. |
If this is the case, what will they do when domains are registered by proxy and there isn't any personal information to be found?
Also, with the price of domains being so low, why can't mom & pop shops own more than one shop? Haven't we all learned, in one form or another, not to put all our eggs in one basket? Are people being punished for diversifying?