| This 31 message thread spans 2 pages: 31 (  2 ) > > || |
|limits of on-page optimization|
is there a glass ceiling?
with google (and other search engines) it seems virtually impossible to really isolate a variable.
however - its increasingly clear to me that on-page optimization has a 'glass ceiling' with google - and it's pretty low.
For example - lets say you have a page with about 200-300 words of content including the keyword in:
First sentence of first paragraph
In most subsequent paragraphs
A few variations (plurals, etc)
Maybe an H2 tag or bold for good measure
At this point - that page has as much on-page optimization as it's ever going to. If it's ranked #20 in the serps - you're probably not going to be able to move much higher via on-page factors.
Toying with keyword density or exact location in the title tag probably won't help. Putting the keywords in italics or an image alt-tag won't do anything to further convince google your page is about a given keyword.
I'm posting this in hope people with either blow my observation out of the water or confirm it...
(it goes without saying there's all sorts of off-page factors that can help)
I am wondering the same thing. I just started a new site that uses these things and has 100 pages. Each page a different subject and should draw hits based on the topic discussed on each page. It's too new for me to relate anything yet but I wanted to know the answers also. I'll be watching this thread to find out.
One thing learned so far. I made a common spelling error on the h1 subheading on the 1st 20 pages. Then caught and corrected it. Google spidered the site while the spelling error was still in place and now I'm picking up SE hits from searches made with the spelling error. I'm adding back the incorrectly spelled word to the key word meta tags but I think I'll need to create a new subset of pages using the wrong spelling for the main site keyword & phrase on each page.
Then link them on the site map so the SE's will index them all. This should bring traffic with a lot less competition since much fewer site owners will be trying this key word to get listed. I should be able to get much higher listing in the results pages if this wrong spelled search term is used. So far it has worked by accident. I guess the SE pro's probably already know this trick.
Recently someone suggested that if pages on your site have had obvious SEO attention (keywords in the title tag, h1 tag, anchor text pointing to the page etc) Google might be penalising the page on searches matching those keywords.
So by putting those keywords in the tags commonly used in SEO, you are effectively telling Google which keywords you want the page to rank on, and making it easy for Google to ensure your site ranks poorly in those SERPs. You would have to resort to adsense.
This could be a paranoid assertion, but then I have noticed recently that in SERPs I follow, the top sites have been badly designed from a traditional SEO perspective.
You hit the nail on the head. There is a selective penalty that Google applies on a page by page basis and only for specific search terms that have been over optimised.
up to the point where the original poster mentioned - you cant call that over-optimization, can you?
Also, should we call it - on-page optimziation in moderation - a penalty? Isnt it just that on-page optimization will help you up to this point, and to rank any further, you need incoming links and PR? The benefits of on-page optimization are lowered across the board I think.
I had some pages hosted on one site with PR4 and few incoming links (no anchor text optimization). Checked the ranking. Low.
Took out the page and hosted it on a site with PR5. No anchor text optimziation. Better ranking.
Took them out and put them on a site with PR6. Better rankings.
All pages were kept one click deep from the homepage.
There's no penalty - remember the 'nigritude' competition results? There's just say, 30% weightage for on page text thats all.
*30% - wild guess
'Fresh content' could have had an effect.
So by putting those keywords in the tags commonly used in SEO, you are effectively telling Google which keywords you want the page to rank on, and making it easy for Google to ensure your site ranks poorly in those SERPs.
not sure to what extent I buy this.
I think it's extremely hard for google to pick out a moderately SEO'd site from a natural one.
I think there is an over-optimization penalty - but it more applies to the extreme cases.
a lot of the stuff in the SERPs -is- painfully SEO'd. If Google really knew how to get rid of it without killing the legit results - they would.
I thought the same thing until last Friday.
I have been at position 9 for my number one phrase (3 words) for months. I tried many variations of my optimization scheme (meta tags/h1/etc.), but my site remained at number 9.
I made a new change on Thursday, and the next morning I was number 4 for the same phrase.
PR nor BL changed.
|I made a new change on Thursday, and the next morning I was number 4 for the same phrase. |
How do you how long it takes for changes to be reflected by google?
The amount of time it takes for changes to be reflected depends on how many sites link to your domain.
>I made a new change on Thursday, and the next morning I was number 4 for the same phrase.
Any chance you can give an indication of the nature of the change? Or even say whether it was to do with the site's structure or a change purely limited to the content on one page? Any advice would be much appreciated!
I added a font size=7 tag inside of my h1 tag (Use .css to maintain proper formatting).
I made that change to the only h1 tag on one page (index.htm). Other than that, I changed nothing.
|Pass the Dutchie|
|The amount of time it takes for changes to be reflected depends on how many sites link to your domain. |
Thatís not entirely true. Its more to do with how often your site is spidered. If you site is indexed in Google has any thing from a few good backlinks, and has fresh content regularly added then under normal circumstances any new alterations will be reflected in a matter of days (assuming that you do not have "nofollow' or 'noindex' or 'revisit after 30 days' in your robot tag.)
I read this idea here a while ago, and yes, it appears to be true, I'd over optimized a page to test this, it dropped to PR 0, I dumped some of the optimization, and it came back, both in PR and the SERPs for the optimized phrase.
If you're trying to rank well in Google focus on "off-page" factors (quality and quantity of backlinks and varied anchor text for those backlinks). With Google, n-page factors account for 25% of SERPS at most. Yahoo and MSN are a different story. On page factors carry much more weight in those two engines.
Thats interesting about using font tags instead of h1 tags. I wonder what the impact on other markup languages like XHTML which do not have font tags is. Perhaps there are some advantages to using HTML over XHTML. Then again, maybe it only penalises H1 tags in HTML documents. I do think it sucks that there is a penalty for adhering to HTML properly, especially as the w3c recommends avoiding font tags etc as part of a wider move to separate presentation from content.
I don't beleive there is a penalty for the kind of optimisation discsussed.
I know a webmaster who runs what are basically hotel affiliate sites - so one of the most competitive areas possible. He does OK, for one country he's in the top 20 and for many regions and towns the top 5.
There are a few things are very interesting about how he does so well (comparitavely).
1. None of his pages have a PR higher than 4
2. He does not actively seek links and has now link exchange page.
3. His site was originally a regional information site and has been around since around 1999. (possibly before).
4. His site is listed in the ODP
Now he thinks he is an SEO wizard. However, I think there are other factors.
As far as SEO goes he always has a top left image with an alt tag with the keywords.
Title has keyword, H1 matches first part of title, H2 matches second part.
Other interesting things include direct links to affiliate listings - so no penalty there.
Infer what you like from this but it strikes me there are 3 major contributors here:
Age of site
Exisitng quality (NOT on topic links)
I do use the h1 tag too, however, I placed the font tag inside of the h1 tag.
My page is XHTML 1.0 - and it's valid, compliant markup.
There is no such thing as over optimisation.
If something is optimised,then it will be set to its optimal level so it will work. "Over optimisation" as some people here call it is just the creation of badly / unoptimised page.
If I put twice the number of h1's, twice the preferred keyword density, etc in my pages, then they are not "over optimised" they are just badly optimised.
Its like saying someone is over good-looking instead of just plain ugly.
|Pass the Dutchie|
gotta agree with ya netnerd.
On page optimisation is just the icing on the cake.
|I read this idea here a while ago, and yes, it appears to be true, I'd over optimized a page to test this, it dropped to PR 0, I dumped some of the optimization, and it came back, both in PR and the SERPs for the optimized phrase. |
I think there has to be other factors as well. Whether or not a page is "over optimized" (if you can call it that as I agree with netnerd) is just part of what controls the results.
Case in point. I too have experimented with some of my sites on adding too much stuff (ie. h1s, bold, italic, text, keyword density, etc) and have also seen the same affect on the sites. Dropping and losing PR.
I have a site that is about 5 years old in which I added all the stuff that would have dropped a newer site, and the older site is #1 for several very competative key phrases.
So I contend that age must have something to do with it also. Keeping that in mind, there are probably many other variables involved and unless you have a wide variety, ages, and types of sites, these types of theories are hard to test.
If you think about it, all the variables can be daunting.
Database (asp,php, cf .net)
All this and I didn't even mention the actual on-page stuff like headers, density, bold, italics, etc. And then you still have all the off-site factors like PR and incoming links, outgoing links, ad naseum, ad naseum.
It gets pretty hard to test under these circumstances. The reason for all this is the G wants it that way. If it was too easy, everybody would be doing it.
Just my 2 cents.
|There is a selective penalty that Google applies on a page by page basis and only for specific search terms that have been over optimised |
Where did you source this information from exactly Piskie?
Yes, and no, call it a rubber ceiling.
Not optimised, adds <b> tags, site goes up in the SERPS by 50%, great.
Optimised, adds <b> tags, site goes up in the SERPS by 5%, ok.
Highly Optimised, adds <b> tags, site goes up in the SERPS by 1%... not so great.
Call it a law of diminishing returns if you wish. The more ticks you've got in the OnPage Factors checklist, the less the next click will help you.
webdude, there were no other changes, only onpage title/h tag, bolding. Plus some in site links also optimized for keyword1 keyword2
I was going for a keyword1-keyword2 phrase, only one.
No other changes took place.
This is an old site, very stable in its performance.
After optimization keyword1 keyword2 dropped totally down in the serps. keyword1 keyword2 keyword3 also suffered, those return 1-4 on average, not many hits, but those vanished too.
After de-optimization keyword1 keyword2 keyword3 came back right away, but keyword1-keyword2 are pretty much gone now[edited, I hadn't checked this in a while].
take it for what it's worth, only one site, but that's what I saw. I used this site because it's so stable in its traffic, pretty much the same week in and week out for a few years.
to bring it back I dropped a few bolds, a few in site link phrases, made sure title and h tags had different text. And I gave up on that approach, didn't want to hurt the site owners further, who are pretty cool.
[Sorry, edited this too:] I hadn't checked this keyword phrase, the latest google updates got rid of the junk, now the main factor seems to be number of inbound links to the sites that are beating it, that makes sense.
I'll buy that. It appears you have hit that fine line. My only point was that I have seen this affect on some of my sites, but not others. This leads me to believe that there has to be some other factors involved. Why some sites and not others? I have some old sites that seem to be able to take a lot more optimization then some of my newer sites. It seems the newer sites are a little more delicate when it comes to optimization. The older sites can stand more changes with no results changing.
Yeah, go figure. I've used this site to expermiment on in the past, usually the stuff has worked, but I did more than I usually do in this optimization, bolds manipulated with css, sitewide navigation links with keyword phrase, index page links with keyword phrase, I threw all the onsite onpage optimization stuff in I could think of, so there was a degree of abuse on my part, but I was curious to see how well google handled it.
<<< hit a fine line
That was my feeling too, just a bit too much, trimmed it and some stuff came back.
Oh, I completely forgot, the page got a PR 0 after the optimization, got its pagerank back after deoptimization and eternal PR update wait, all other pages of site had and have PR. I spaced that, that's why I knew for a fact that a penalty had been triggered and applied, sorry, that's a big bit I forgot to mention.
We'd really like to get in the top 10-20 for this phrase, but there's something else going on that I need to figure out, if it just comes down to sheer quantity of inbound links there's not a lot I can do.
They're number one for their niche 'widgetx' keyword, but they'd like the more generic 'general widget' phrase, since that has so many more searches for it, and they are definitely a key, well known among those who know, player in the general widget scene.
On page still counts a bit, however your not going to on-page yourself to the top of any commercially competitive serp.
As has already been stated in this thread, concetrate on off page.
rich42 - I recommend putting some thought into site structure... what's targeted where... as well as into internal linking, which can help both originating and destination pages.
Tuning a page as you get inbounds is also helpful... text proximity, eg, onpage and in titles. You can also look at your log files and at rankings, and see where you can enhance rankings of a page for secondary phrases.
You can't always predict how searchers will search, so adjusting existing pages (and/or building new ones) to catch variations will increase traffic.
For a helpful discussion of site structure in relation to targeting and internal PageRank distribution, take a look at this thread...
Search Engine Theme Pyramids and Google
Optimising the Pyramid for PageRank
*the page got a PR 0 after the optimization, got its pagerank back after deoptimization*
That to me is an oddity, if it was the main page, it might have been unrelated.
Just curious to know if any of the tests made here that resulted in drops in SERPS also happen to show older cache dates, or did the newest test results always have newer cache dates.
| This 31 message thread spans 2 pages: 31 (  2 ) > > |