Forum Moderators: Robert Charlton & goodroi
<edit> Suppose you change the navigation on your site so that all the home page links say "keyword keyword home" instead of "home". Will you incur an over optimization penalty?
My site appears to be suffering from this penalty. What are the views of the SEO experts here at WW? </edit>
[edited by: lawman at 3:00 pm (utc) on Dec. 6, 2005]
[edit reason] Edited To Conform To TOS 10 [webmasterworld.com] [/edit]
It's good to have a word on a page if you want to be found for searches for that word. It isn't "over optimization" if you repeat the word a million times. That is just rotten optimization.
Optimizing a site is just to put it in its most favorable light. That's not "semantics". That is the point.
It doesn't make sense
Why not?
You are of course entitled to your opinion but dismissing the arguments of others without providing any evidence to the contrary is not helpful Steve. Are you trying to say that many spammers websites don't follow the model I described? Search engines look for signs of spam. OO is a strong sign of this and that is why an OOP makes perfect sense.
It isn't "over optimization" if you repeat the word a million times. That is just rotten optimization.
Why be pedantic? OOP or "rotten optimisation ", what's the bl**dy difference? I am happy to refer to this as OOP since that is the term that has been in use in this community for more than two years. Most of us know that just about everything that affects the position of your site is just part of the overall algo but it makes no sense trying observe the algo's effects as a whole. We must dissect it and analyse some of its more radical and specific effects. None of us would get anywhere without doing this.
Many of us who have suffered from it know the effect we are talking about. Try a Google search on OOP SEO and you will find 127,000 pages about it ... but then maybe all of these people are talking nonsense too. ;)
This is in direct contrast to Yahoo, where you have very high quality sites, tripping an over optimization filter and then being removed from the index. Their algorithm and the approach to dealing with what they consider to be sites outside of their guidelines is crude at best. Often over optimized sites rank well and quality sites are given the death penalty. Their results often present as extremely bizarre.
If you have extreme aspects to your site, i.e. optimization tactics taken to a point that is above and beyond any reasonable limit, you have a bad site. Google handles this as it should be handled, you rank poorly. Yahoo, in very stark contrast just does not have the capability to approach it in this manner, which really solves the fundamental problem, ranking sites where they belong.
Is there an OOP? Yes, but not where you might think.
but this is steveb's thing, if he isn't interested in reading other people's examples of empirical evidence and prefers his own opinion, that's his right. Personally, I'm interested in other people's experiences since then I might learn from them, but that's just me.
But just for fun, let's say he's right, it's happened before, so let's not discount it.
If he's right, that means you can never get penalized for overoptimizing your web pages....
hopefully at least some of you will start scratching your heads right about now, and you'll say: wait a minute, are you saying that I can do ANYTHING I feel like, put as many keywords, pump up keyword densities to 97%, insite link text spam to my hearts content, create <b>keywords</b> with css for b{font-weight:normal;}, and so on, and I won't get penalized, ever? Amazing, the things we can learn here.
So what are we all waiting for? Let's get started. Oh wait, I tried this, and got penalized. Ok, you guys do it, I'll watch.
I think word frequency is not important and certainly not penalised, as many 'non seo' written pages will repeat a keyword many times for perfectly natural and legitimate reasons. It would be pointless for google to penalise when often word repetition is legitimate, therefore the solution is to ignore.
Therefore, doing too many seo tactics probably results in having no negative effect. HOWEVER, I think Google looks for seo activity which, if several tactics are done on a page, raises a flag. For example, a specific phrase being targeted in the title tag, h1 and anchor text is suspicious but not necesarily damning. Therefore the normal plus points given for these are perhaps reduced or at worst not allocated and the page has to rank according to its other qualities, such as 'links in', word proximity and pr etc. The result is that the more you optimise for a phrase using standard tactics, the less 'trust' google will have in your page and will want to see other qualities before it will rank you. In this way, if the 'repetition' is legitimate, other qualities will confirm the relevancy and allow you to rank well. The algo could be quite simple, something like "if keyword/phrase is in title/anchor/h1 then ignore unless x/y/z is also present"
This would suggest there is no 'OOP penalty' but instead an 'OOP filter' which demands more signals of relevance and quality before the page will rank. In this way, natural text which appears 'over optimised' or 'poorly optimised' but in fact is quite natural can still rank well.
The word "optimisation" causes all the confusion here, because in theory you cannot do anything but achieve 'optimisation' or fail to achieve 'optimisation'. I think many seo's now use the word as a collective noun for a series of tactics and if you do too many you can trip a filter and thus 'over optimise'.
Perhaps OOP should stand for "overuse on page" and refer not to a penalty but a devaluation of the target phrase.
In short, prety much yes. As long as people aggressively miss the entire point of SEO, they will come up with terrible, foolish conclusions.
The point of optimizing is to get things "right", and "right" happens to be a moving target. There is no "over right", there is only "not right".
Optimizing is not rocket science. You are just making a pot of soup into which you mix ingredients. Adding pepper can be good, but adding two pounds of pepper is not "overoptimizing", it's just stupid. Likewise, while a banana might be good in one meal, and curry might be good in another, combining a banana with curry will create an unsatisfying result. Once again, you didn't overoptimize the food, you just made a bad combination.
People who think putting strawberries and ball bearings into a pot of chili will overoptimize the chili need to learn more about cooking, not complain they are suffering from a penalty.
There is no "over right", there is only "not right".
Who said anything about over right? You are really good at semantics Steveb but I believe that you are totally wrong about this. Adding some pepper is seasoning, adding two pounds of pepper is stupid but it also "over seasoning" in my book.
What is more interesting is to what degree, if any, "over optimisation" does actually result in a "penalty" and which "ingredients" (or mix of) are the crucial ones. I have seen no evidence that using the same keyphrase in the ingredients of page URL, title, H1, and anchor text causes a "Google penalty" nor would it make any sense if it did. Duplicate content and link structures are maybe areas more worth exploring.
The reason why pages rank or don't rank would be more complex than "over optimisation". So I tend to agree with steveb.
That of course is the main thing. All those things are part of optimizing the site, and obviously there isn't any evidence of this being a negative.
"Over right", "over best", "over optimize", just stop focusing on such concepts. They don't exist, whatever you call them. Its probably unachievable, but an optimized page is simply a page presented in its best light. Optimization is the constant effort to approach that perfect level of optimization. You are trying to score a 100. You can't score more than that. Adding a bit of pepper may take you from a 90 to a 93, while adding two more pounds of pepper might take you from 93 to 24. Adding the two pounds is not the "right" choice. It was the wrong choice.
"Over right", "over best", "over optimize", just stop focusing on such concepts. They don't exist
They do exist because we do not know the criteria of the optimization. Google does not tell us what triggers the filters.
We simply cannot do the things right if we do not know what the right things are. This is the problem.
So we try and see (that is all that we can do): better, better…. Oops we are nowhere. We have over optimized.
The problem is that the over optimization Google filter to my opinion is too severe. It violates their "do not be evil" motto. They loose really outstanding sites in their results and preserve the average and dull sites. I would have rather allow more spam than loose the outstanding and interesting sites.
So the problem is to guess what Google consider right. Just to know where to stop in the optimization.
Vadim.
There is no sandbox. There is no OOP.
What there are, are multiple algo's, and a variety of filters associated with each. I think.
IF you use the same tactic too much, or the same kw too much, and/or in too many ways, then yes, you are at risk of being filtered out for ranking on that kw. We've proven it over and over again to our satisfaction.
In it's simplest form:
Try running the same three word backlink anchor text on 80% + of your backlinks (if you have that sort of control). Do you know what'll happen? If you DO have that sort of control, then you'll probably stop ranking well for that kw (in this case, there are probably enough other factors in your control, also leaning the same way, that you're pushing too far).
OTOH, if you DON'T have that sort of control, it's likely that you're running a 'cleaner' site and don't have all your SEO techniques leaning the same way, and you may just get away with 80% of your anchor text being identical, because enough other elements are "natural"; though of course in this case, you don't have that sort of control anyway, so it's not possible for you to make 80% of your anchor text say the same thing, and you're more likely to be fine.
Ironic, isn't it? ;-)
There is no sandbox. There is no OOP.What there are, are multiple algo's, and a variety of filters associated with each. I think.
"There is no duplicate content filter - it's just part of the algo".
"There is no sandbox - it's just part of the algo".
"There is no over opimisation filter - it's just part of the algo".
"There is no link structure filter - it's just part of the algo".
If we are getting into semantics again I don't think there are multiple algos. There is only one ranking algo which uses many different elements in its decision making process. However it is far too complex to be treated as an entity. This is why most of us are happy to use terms like sandbox and OOP in reference to these individual elements. Those of you who are clever enough to analyse the algo holistically should perhaps leave now and let we lesser mortals get on with our murmurings.
"I have seen no evidence that using the same keyphrase in the ingredients of page URL, title, H1, and anchor text causes a "Google penalty" nor would it make any sense if it did."
As 2by4 said earlier does this mean that we can now do what we like in this respect? I don't think so and I believe that on one of my sites I have evidence to the contrary. On this site I believe that I Over Optimised and that I am now paying the penalty with thirteen months in the sandbox.
(Oh sorry - I forgot - there is no sandbox). ;)
This has caused sites that purchased one-time optimizations from us years ago to shoot up in the rankings. Conversely, sites that we adjust meta tags and keyword density a couple times per year seemed to have been hurt in the update.
Is anyone else seeing what looks like a benefit to stale pages and a decline in rankings for pages that have a more recent date modified?
My theory being that Google may be saying if your title tag and description changed 6 times in the last year and the rest of the page is similar you may be over-optimizting the site. (Sorry steveb, im using that term)
What good is that to a user? It's the navigation!
Perhaps this added an extra element to my pages that tipped it over the cliff.
Bye bye 50% of my traffic! Hello workload!
Hello workload!
Without seeing the site I'd maybe consider the first part of the workload as replacing those JavaScript menus with CSS-suckerfish type dropdown menus that Googlebot can actually crawl (unless you have the navigation replicated somewhere in HTML).
It's replicated in HTML, for sure!
What's a css sucker fish menu?
I'm getting too old for this constant adaptaption!
I have to disagree, playing with words does not equal being good at semantics (the study of meaning), it's just playing with words, even when it makes you say things you couldn't possibly really believe, at least not if you have paying clients. Personally, word games kind of bore me, although I guess they're fun if you have nothing better to do, or if they let you ignore something else not that great in your life, whatever, I'll try not to speculate further in that direction.
I've found combinations of actions that will result in penalizing the targetted page, similar to what caveman described above, but localized to the site itself. I now avoid those tactics, and do not get my pages penalized.
If you're reading this thread, you can follow some very bad advice, and believe that you cannot trigger a spam filter by your on page and in site methods, but don't look for sympathy when you find that you did in fact trigger this or that filter. I have seen these methods for at least one year on other forums I frequent, and they are real, and they will consistently trigger penailties, sometimes sitewide penalties. In fact, I'd say at least half of amateur mistakes I saw during that time were direct results of over optimizing the onpage code.
It's totally irrevelevant what you call these combinations of algo components, although for practical communications purposes the members of the seo fraternity give them names to make the various behaviors easier to discuss.
This is a useful tool, it's like shorthand... no, wait, it's like language itself, we create words to be able to discuss and control our environment. Then if we get really bored we argue about whether or not the referent to that word exists or not.
As beedee says, of course it doesn't exist, it's all the algo, filters, penalties, etc. But it's boring, and not very useful for us to say, oh, the algo is behaving in this way now, describing all those behaviors in each posting, it's more efficient to just give that behavior a name, it's why brett started naming updates for example. It's just easier to say 'bourbon' instead of describing the new observed behaviors, in a long list, of the google algos.
Jagger even more so, since it was 3 separate things. But Jagger doesn't exist, OOP doesn't exist, the sandbox doesn't exist. Of course it doesn't exist, no one can point to it, no one can see it, these things really don't exist. But we can still use names to describe collections of behaviors.
Of course, when people move from denying the existence of the name itself to denying the existence of the behaviors or collection of behaviors the name serves as a label for, it's kind of... odd. Like why bother wasting the electronic transmission of those ascii characters?
anyway, not much is gained by arguing things that you know are issues to watch out for from direct personal experience, I think there's probably better ways to spend time.
why don`t we all just agree that no one knows for sure
Some people know for sure, because they take the time to run specific tests. Then there are people who know for sure, but they have their own agendas and make posts to the contrary to keep the masses dazed and confused.
