| 5:21 pm on Jul 8, 2005 (gmt 0)|
Links change, and Google shows us only a small proportion, but it does look like Talar's looking at a site that doesn't have the words in the page or in links to it.
If it doesn't, and didn't, and if the ranking isn't due to a parsing error causing text before a link to count in the link, then two possible causes would be:
a) Google is using the words near the incoming links.
b) Google is using other words on the pages with the incoming links. (i.e. "theming")
| 5:25 pm on Jul 8, 2005 (gmt 0)|
Two possible explanations:
1. Google has instituted a "too relevant" filter. G does not understand that legit publishers use keywords in titles, text and URLs and believes that only a spammer would use the word "widget" to describe a page about widgets.
2. Scapers have so severely poisoned domains which used to rank highly (due to use of keywords in tiltle, text and URL) that most of those sites have now been seriously downgraded in the G index, leaving only marginally relevant results to appear.
[edited by: Atticus at 5:37 pm (utc) on July 8, 2005]
| 5:36 pm on Jul 8, 2005 (gmt 0)|
The links pointing to the page could be within the website, even within an image link (text located in the alt text). Don't forget to check those.
Also might want to check backlinks using Yahoo. I wonder if the text in the title tags, too.
| 6:11 pm on Jul 8, 2005 (gmt 0)|
|Since May/20 I’ve noticed a significant degradation of Google results. Here is just one example however since we can’t post specifics I will just use “abc xyz” as the keywords. |
We ALL have noticed that! That's the topic of many threads around here (all now closed). Latest thread was here: [webmasterworld.com...]
|If you search for “abc xyz” the first result has none of the keywords. To find a clue to the problem I click on “Cached” and it reads, “These terms only appear in links pointing to this page”. However none of the pages linking to that webpage seem to have the keywords either. |
.....I don’t want to accuse the site of spamming since it doesn’t seem to benefit from the position because “abc xyz” is a very specific phrase regarding to finance and the website is completely unrelated and it has no ads or links that could be of benefit.
That is EXACTLY what I and many others have been seeing, except in my cases it's most of the results on the first pages that don't have the searched-for words, and not to mention are rarely relevant to the search! G doesn't want relevancy, they've proved that. Common sense tells you that when you search for something, that THAT is what you want to find! G doesn't agree. They are placing more and more precedence solely upon how OTHERS link TO you which is NOT the way to go. Content should be what matters, the content on a webpage, but it doesn't. You're chained to how many others link to you, and how they do their linking.
|Any idea of why this is happening? |
Yep, general consensus is G is screwed-up. "Atticus" summed it up perfectly: if you sell "blue widgets" you're not allowed to say on your webpage "blue widgets" more than once! Oh noooooooo, if you did, then it would be "too relevant" to your search! Senseless.
I think the best thing G could do (other than give it up) is to put a link next to each result in their SERP's that said "Non relevant?" then you could click that for all the countless BS results and that data would then be sent back to G so they could investigate the complaint* of non-relevancy (*so a competitor could not do it and hurt you). But, that's too easy and would work too well for them. They'd rather let their algo screw things up and harm legit innocent sites.
| 8:37 am on Jul 9, 2005 (gmt 0)|
|Yep, general consensus is G is screwed-up. "Atticus" summed it up perfectly: if you sell "blue widgets" you're not allowed to say on your webpage "blue widgets" more than once! Oh noooooooo, if you did, then it would be "too relevant" to your search! |
Clint, I appreciate that your current Google experience is not good but I don't think that Atticus's summing up is correct. I have good content sites that are still optimised the way they have always been. One of them does extremely well for a four letter acronym that is repeated all over the page in both content and anchor text. Others do quite well and one has bombed as a result of Bourbon (I think!). I would say this is about par for the course in any update.
I don't think that we can base our findings on one spurious result, which we appear to trying to do in this example. If a few others come on board this thread telling me that they too are seeing "no keyword" results then perhaps I will reconsider :o)
|1. Google has instituted a "too relevant" filter. G does not understand that legit publishers use keywords in titles, text and URLs and believes that only a spammer would use the word "widget" to describe a page about widgets. |
Atticus this is a bit of an exaggeration but the possibility of a "too relevant" or over optimization penalty has been discussed at length in here over the last year or more. This is one recent thread on this subject [webmasterworld.com...]
| 10:14 am on Jul 9, 2005 (gmt 0)|
I think that you might find that the linking pages not listed by Google in their link examples would have the phrase
The other possibility is that some of the linking back pages may use dynamic content that when Google botted had those keywords, however they have updated with a fresher page that doesnt have the phrase anymore - different time periods for linkbacks than the fresh cache pages
| 10:18 am on Jul 9, 2005 (gmt 0)|
|I think that you might find that the linking pages not listed by Google in their link examples would have the phrase |
Agreed. This sounds like the best explanation.
Wonderful new verb :)
| 3:23 pm on Jul 9, 2005 (gmt 0)|
G doesn't do anything across the board. The "too relevant" filter (intentionally implemented or not) does exist in some industries or for some sites.
I manage some sites that still rank for specific keyword combos. But my sites which which used to draw the most traffic have all but disappeared from the SERPs. Searches for unique phrases from these sites in quotes usually do not bring up my sites, but do list a number of scrapers with my snippets.
I agree that sites may be ranking well based on inbound links which G does not display in the back links. But that doesn't sound particularly relevant to me. My sites list specific phrases in text, titles and URLs and have hundred of inbound links from edu and k-12 sites, yet are nowhere to be found.
Do some searches on various KW combos and I have no doubt that you will find some areas where the top results consistently fail to list sites with the KW in the title or on the page.
It is real, but not universally applied.
| 7:04 pm on Jul 9, 2005 (gmt 0)|
|I think that you might find that the linking pages not listed by Google in their link examples would have the phrase |
Unlikely. The webpage is completely unrelated and has no banners or any other advantage to its webmaster. It is as if you searched for [forum for webmasters specializing in readability] and the first result was a site about soccer.
|The other possibility is that some of the linking back pages may use dynamic content that when Google botted had those keywords, however they have updated with a fresher page that doesnt have the phrase anymore - different time periods for linkbacks than the fresh cache pages |
Unlikely since it has been there too long for this to happen.
[edited by: ciml at 7:19 pm (utc) on July 9, 2005]
[edit reason] No Solicitations. [/edit]
| 11:20 am on Jul 10, 2005 (gmt 0)|
Talar - I assume you have checked all of the pages linking to this site (not just what Google shows as links, but a complete listing from Yahoo etc) and have taken into account that the backlink pages may be the authors own "other sites"?
I know of a huge amount of sites that use completely off-topic sites within their own networks to get ranked and then as part of their gameplan, after jumping to the top 10, add a good amount of content to rank for anything they want.
It doesnt matter if the receiving page has anything to do with the phrases and never has (although granted it does help)- you can rank for widgets as long as you have a page that has any content (no need to be about widgets) and lots of links about widgets coming in - thats just a fact of the way Google has always worked.
| 11:42 am on Jul 10, 2005 (gmt 0)|
This all depends on how competitive that term was, Talar? How many other results were presented?
If it was a competitive term then it would seem weird that it has the #1 spot, even if there is a lost link out there somewhere using those keywords. If it's an obscure term then that wouldn't seem to dramatical as evidence that Google are doing this deliberately in order to make people cough up for AdWords (which is my personal belief though).
In trying to be unbiased - I have found that I am having to dig deeper to find things than say 3 years ago. Just the other day I was reinstalling a thing on linux, so I searched for instructions using the same keywords I used to find instructions when I first installed this thing some years back. Back then I remember finding instructions easily, but this time, most of the stuff that came up was useless and I spent about 10 minutes looking for something decent - on page 12 I do believe it was (found something useful on page 1 i believe the first time).
| 12:30 am on Jul 14, 2005 (gmt 0)|
|I assume you have checked all of the pages linking to this site |
No, it would take forever to do that. In many cases links can explain the position, however in this case it would make no sense even if it were spam since there is no benefit to the webmaster. Also, s/he didn’t even put a single of the keywords on the webpage. All subsequent results are highly relevant from banks and financial companies.
Apparently you are correct, this is how Google works and their results are getting less relevant all the time.
| 12:32 am on Jul 14, 2005 (gmt 0)|
|How many other results were presented? |
83,000 results in total. Is that considered obscure?
| 12:56 am on Jul 14, 2005 (gmt 0)|
|83,000 results in total. Is that considered obscure? |
That borders on microscopic in the area of pages returned. A couple of left over text links from well ranked 'authority' sites could probably dominate this search.
Is it possible that the domain used to have another topic on it and some of the inbound links still exist, or exist in the G index? If it is an old domain, this is not only possible, but likely.
Remember, if a page does not change, no body recaches it.
IOW If a page stays the same (last modified date and/or e-tag) there is no reason for a search engine to reload the page, nothing has changed, so the current cache is correct and accurate.
This means that if sites/pages have not been updated in extended periods, but still exist (even from 3 or 4 years ago), the links are still 'there' and the text is still the same, giving the linked to site, well aged backlinks about a topic.
I have actually seen pages that are 3 years old (not updated) well ranked for terms, because they are never revisited to see the 40+ broken links on the page since they are never updated... and if they have not changed, the cache is still acurate.
My guess is somewhere out there are a couple/few well aged backlinks about the other topic from pages that are long forgotten about, but still live and since the pages have not been updated, they will remain.
| 6:00 pm on Jul 14, 2005 (gmt 0)|
|If a page stays the same (last modified date and/or e-tag) there is no reason for a search engine to reload the page, nothing has changed, so the current cache is correct and accurate |
However, as indicated, there is not a single of the keywords on the webpage. If we assume that long ago it was about the topic (had some of the keywords), it no longer does (and hasn’t had them for a long time) so that webpage had to have changed and re-cached, otherwise the current cache would not be correct/accurate.
By the way, the result for that webpage doesn’t have the “cached” link. Why is that?
Just to add more information. The keyword combination has 4 words so I picked just 2 of them and did a search. There are 23 million results and that website still ranks among the top results (although not the 1st one anymore). All the results I reviewed under that one are more relevant and contain the keywords.
| 6:17 pm on Jul 14, 2005 (gmt 0)|
Those links without title or description are the LINKs to the page I think.
Reasoning: I have a page that has noindex, nofollow in the page. However 3 entries are showing for it, they correspond exactly to the 3 different links to that page.
| 2:11 am on Jul 15, 2005 (gmt 0)|
|Do some searches on various KW combos and I have no doubt that you will find some areas where the top results consistently fail to list sites with the KW in the title or on the page. |
Many of the examples I have seen like this are cloaked pages.
Few things are really absolutes with Google.