8 year problem with disappearing rankings - why?

[SilverSpirit]

I am at a loss to understand why one of my websites is constantly disappearing for its major search phrase [green widgets] on Google ever since it was created 8 years ago.

Sometimes, it manages to get to the third page of results; but most of the time it is nowhere to be found. The site has never been banned, and can always be found for more specific search terms.

This is only one of my sites, and most others are doing just fine, and appear on the first page for their main 2-word search phrase. I am doing all by the book, no cheating whatsoever, and my meta tags and code are very clean.

[green widgets] is a competitive search term.

Here is the result of my systematic research about the first 10 results that appear on Google for [green widgets]:

9/10 sites have more pages than mine (mine has 76).

2/10 sites only have more inbound links (I have 4,218).

6/10 sites do not have a listing in Yahoo directory (mine does have a Yahoo listing).

3/10 sites do not have a listing in DMOZ (mine does have a DMOZ listing).

3/10 sites have more outbound links (my site has sitewide footer links to 7 sites part of my network; all about [widgets]).

6/10 sites have more internal navigation links.

4/10 sites have a higher occurrence of the word [green].

5/10 sites have a higher occurrence of the word [widgets].

9/10 sites have a lower occurrence of the word [widget].

7/10 sites have the word [green] in the url (and so does mine).

7/10 sites have the word [widgets] in the url (and so does mine).

8/10 sites have more words than mine on the page displayed (mine has 846 words on its index page).

9/10 sites have been more plagiated than mine according to Copyscape.

2/10 sites have a higher PR than mine. 4/10 have a PR lower than mine.

1/10 sites uses hidden text

The most striking part of the above is that 9/10 sites that are doing well have more pages than mine, and their index page has more words than mine. Is that really what I should do: write more pages, and add more words, in order to appear among the first 10 results?

[SEOPTI]

This is the long term strategy Google prefers, they force people into Adwords by making their rankings appear and disappear or rotate, call it whatever you want ...

[martinibuster]

You didn't check the words that your page contains that the pages listed in the top ten, top five, and bottom five do not have. I've seen before where certain words were keeping a site from being relevant for shorter tail phrases. After removing certain words the site jumped up in the rankings...

[tedster]

Even though many people hold the "force business to use Adwords" opinion, I can't buy it. I just don't see the evidence, and I look at a LOT of data.

Yes, when good rankings go away, then Adwords may be the response that some businesses make. But that's not the same as saying Google is creating this effect intentionally. If such a strategy were ever proven it would be a terrible long-term problem for Google, and Google as a whole is much too savvy, even cunning, to commit such a blunder.

And before I forget my manners, welcome to the forums, SilverSpirit. As I understand your situation, this particular site never seems to get above page three for it's main target phrase. There are some factors that your opening post doesn't get into that may be in play.

1. Quality of inbound links, not just the quantity. For example, how many are reciprocal linking arrangements? How many are freely given "editorial votes" from the linking site? How many links are in the actual content area of a page, rather than the footer link or even just on a page links?

2. Outbound links both on your competitor's sites, and yours. Have any of the link targets ever been sold ro changed, and turned into bad neighborhoods?

3. Anchor text, both for internal links and inbound backlinks.

The above factors are worth some deeper study. Also, look at how many backlinks point to internal pages rather than the home page. How many total links are on the Home Page? Is the site structure well designed, especially the main menu?

And though I appreciate your intentions not to "cheat", what exactly does that mean to you? Google has certainly shifted the line in the sand over their history - with link selling being a big shift in more recent times.

How competitive is this two word search? Not only how many total results, but how strong is the presence of Adwords on that SERP - that's a strong signal for how heavily monetized the kw is.

Do you offer unique and valuable content, or is it commonly available on other sites as well? Does your site have any technical canonical problems [webmasterworld.com]?

The above issues, and more, are discussed in the Hot Topics area [webmasterworld.com], which is always pinned to the top of this forum's index page. Reading in there may give you even more ideas for directions to pursue.

[edited by: tedster at 3:57 am (utc) on Jan. 28, 2009]

[tedster]

Martinibuster posted while I was typing and put a finger on a big factor that I've also experienced. In fact, I had a similar challenge once with a client who could not get above page 3, even though they were only one of two companies in the world that created the product (also a 2-word phrase).

It was paying attention to "term co-occurrence" that met the challenge and they now have a stable position #2 and #3. In short, the copy they had was too restricted to their specific keywords and did not have ANY of the kinds of semantically related words that would be natural.

We don't discuss specific keywords here, for the sake of clarity about this topic, I'll offer a purely hypothetical example. A page about apples might very well be expected of have at least some (but not ever all) of the words "fruit, juice, cider, orchard, harvest, tree, green, red, ripe, stem, core, seeds, Gala, Delicious, Fuji".

[SilverSpirit]

Many thanks, tedster and martinibuster for the helpful replies.

You didn't check the words that your page contains that the pages listed in the top ten (...) do not have.

Excellent point, thanks. Will do. My site is in fact about [special green widgets], so one could think that the word [special] is a problem; however other sites in the top 10 also have that word. Will look for other words that may make my site appear too focused on [special] or other particular widgets, and that may prevent it from being in the top 10 for [green widgets].

1. Quality of inbound links, not just the quantity. For example, how many are reciprocal linking arrangements? How many are freely given "editorial votes" from the linking site? How many links are in the actual content area of a page, rather than the footer link or even just on a page links?

My inbound links are much better quality than the other sites': more relevant and with natural anchor text (not 4,218 times the same anchor text). I have hardly any reciprocal links.

Outbound links both on your competitor's sites, and yours. Have any of the link targets ever been sold ro changed, and turned into bad neighborhoods?

I can vouch for all the links from my site. I check all my links regularly.

Also, look at how many backlinks point to internal pages rather than the home page.

I have 2,860 inbound links to my homepage and 1,358 inbound links from other websites to my internal pages. Compared with the top 10 websites, this proportion is not significant (some websites have 100% links to their page that comes up for [green widgets]; some only 50%.

And though I appreciate your intentions not to "cheat", what exactly does that mean to you?

No hidden text, no link selling, no link schemes, perfect html.

How competitive is this two word search?

[green widgets] gives 623,000 results and plenty of Adwords. It is quite competitive.

Do you offer unique and valuable content, or is it commonly available on other sites as well?

My content is unique (proof is that Copyscape does not find any plagiarism) and valuable... at least to those who have an interest in [green widgets].

Does your site have any technical canonical problems?

I am quite sure it does not. My .htaccess file has a 301 redirect from non-www to the www version.

Many thanks for the example about apples and "fruit, juice, cider, orchard, harvest, tree, green, red, ripe, stem, core, seeds, Gala, Delicious, Fuji" - will see how it can be applied to my site.

Interesting that no one has seconded my idea that the amount of pages and words on my site has anything to do with its poor performance.

[Shaddows]

If you get more pages, typically you need more links to get them ranked. It depends on the quality of those pages as to whether they help your 'focus' pages without links (i.e. they don't rank, just help your other pages rank).

Simplistically, every page has a Goog-given 'vote'. This gets modified by PR. If you had enough NON-SPAMMY pages, all those extra votes would tot up.

I doubt very much word count has an effect (except to increase the content-to-code ratio), unless your pages are extremely light on text in the first place.

[aristotle]

Have you ever heard of the over-optimization penalty? I've read that some people think it's a myth, but apparently some also think it's real. If it does exist, it would apply to a situation that's "too perfect". In other words, your site (and main landing page) could be over-focused on and over-optimized for "green widgets". I'm no expert, but I have heard of this, and just wanted to mention it as a possibility.

[Shaddows]

OOP is usually down to keyword stuffed H1, title, meta and alts. Additionally, it could be unvaried anchor text on inbounds, or unnatural frequency in the text (especially without expected co-occuring phrases).

[SilverSpirit]

Thanks for the ideas. I'm pretty sure my site in not over-optimized, and inbound links definitely have varying text. I have also checked "term co-occurrence", as suggested by tedster, and I am reasonably sure that no crucial terms are missing on my site.

I have pursued martinibuster's suggestion about reject words (words that may cause Google to reject my site towards the end of results) and came up with a potentially interesting discovery:

My site is about green widgets, which we love. These widgets are beautiful and charming; their charm comes from the personal attention we give to each widget, as green widgets are our beloved products. Each widget matches exactly your needs.

Our site isn't about matchmaking at all, and does not have any adult content... but could the words love, charming, charm, personal, beloved, matches suggest otherwise to Google? - Anyway, since none of the top 10 sites have these words, I have removed them now as a precaution. Wait and see.