Summary of SEO - What and What Not (My beliefs on SEO)

[FredJones]

Hello Warriors,

In spite of every single discussion that gets done on SEO, I still see literally hundreds of discussions on a number of topics that ask, "exactly what to do for SEO".

And I see the most confused ones going in circles, wasting efforts and money, and running pillar to post are newbies, and some intermediates too.

So let me talk a little on SEO for Google. I hope it would remove many confusions once and for all.

But who am I to talk on SEO? Please let me cite some of my credentials and achievements.

I have been on the first page and #1 rank for Google in many terms, including competitve ones that I choose to. I have been ranked #1 for terms that I have targeted, with a single sale leading to more than $200 in commissions even for physical products (yes, products priced even more than $3,500). With a fair bit of conifdence I can claim that I have seen the SEO field to a good extent.

And my education being on Computer Science all through (have a Masters of Technology degree and hope to do a PhD also from some world-leading university in near future) I have a rare combination of understanding the computer science behind it, which is critical and goes missing in most of the discussions.

EDIT: I decided to add this line now after reading Mike's post below - whatever I describe here is purely on my understanding but with the best of intentions and with the best of my capabilities. But anyone reading this is suggested to take action only as per their belief and understanding. Please do not trust me if you do not want to - and if you trust me then please do that at your own responsibility. This is purely an informational thread and no selling is intended - please do not buy any SEO related or any other product from me anywhere, here on Warrior Forum or anywhere outside, beacause of this thread. End of EDIT.

Lets go ahead to the topic now. Technically I could go much deeper, but I want the thread to be simply enough so that even newbies with no related background can easily understand it. So I am avoiding all the mathematics and writing in simple English here.

There are two main sides when it comes to SEO on Google in practice (I am not associated with Google in any way, and what I am saying here is what I have seen to work repeatedly, and some general theory including the related Computer Science portion). (1) On page SEO and (2) Off page SEO. To see top rankings in your SEO efforts, you have to shine on both the fronts.

Important: in my belief, if your on-page SEO score is A and offpage SEO score is B, then your total SEO score is (A x B) and not (A + B). What does that mean? That means, a slight deviation in any of the two compoments would hurt your score badly.

Next fundamental question would be, how to do on-page and off-page SEO?

How to do on-page SEO?

On-page SEO is a combination of the following factors:

Page-title
Meta-tags
Meta-description
H1, H2 etc tags
Content
Boldened, italicized and underlined content
Outbound (relevant) links

There are differences of opinions on whether some of these parameters are counted any more, such as meta-tags and meta-description and whether or not outbound (relevant) links hurt more than help, but each of them do set up the context for the search engine. And you must write your on-page content accordingly.

Now comes the keyword. The main keyword must appear a few times, and there should be related keywords (use Google Adwords Keyword External Tool to find related keywords easily in a minute or so and for free). And the overall topic must be relevant - the stage of Natural Language Processing is advancedd enough nowadays to differentiate between obvious bad page versus good page (not yet good enough to identify a well-spun article with its original one, but may catch badly-spun ones as garbages).

When it comes to the keyword (niche), there is a mixture between an off-page factor - your competition - with the on-page factor. This is the TF-IDF (stands for Term Frequency - Inverted Document Frequency) quotient.

Inverted Document Frequency (IDF) is computed from what we call "competition" in practice. The higher the competition is for a keyword, the lower is the IDF score. It is like inverting the number of occurrences of the keyword in the Internet to see how scarce/available the term is. So if some term like "the", "is" etc appear in every document, the appearance count is very very high, so th inversion (1 divided by this count) is extremely low, and the term is practically ignored - automatically. If you realize, IDF is the off-page factor that we look at when many of us start selecting a niche hoping to do SEO on it later.

Term Frequency (TF) is a different beast. This is the what we call the "keyword density" - how many times does this appear on the page. Now, every LSI word has a similarity score with the main term, so every LSI occurrence improves the TF count without touching the term. The way I think Google computes keyword stuffing is by looking at TF (sufficient occurences but within a higher bound - and the longer the article/content is probably the lower the TF should be in terms of percentage), the score gained by the term due to LSI relevance (frequency of each LSI applied x the relevance score of that LSI), and adding this TF with the LSI-related score to come up with the TF-score.

Now that it has got the TF-score and the IDF-score, it will multiply these two to come up with the TF-IDF score.

By the way, all the factors such as title, meta, content, H1-H2-H3-etc andd bold-italics-underline etc mentioned above, get merged to form parts of the TF-score. So you now know what to do to improve each of these factors now and get a good TF score - apply the core keyword not too many times but definitely do apply it, and then follow that up by mixing a good variation of LSI (Latent Semantic Index keywords - which are relate and similar keywords with respect to the main keywords).

Getting a good TF-score is critical once you have selected your niche and keywords (and thus your IDF score is frozen) to get the best of on-page SEO.

Do you now know why some article directories, for example, rank better and easier with the same/similar article/content than others, even if they appear to have the same page rank?

That is because, they provide a great framework for the on-page SEO to happen automatically (for example, the right kind of inherent tags get generated capturing the keywords and LSI keywords).

And with some tweak to your webpage, there is no reason you should not be able to do the same.

Also, by using the right keyword combined with the right LSI, different articles will rank differently in search results for the term even if they are in the same article irectory, for example. Or, the rank in search results of your page will change.

Now of to off-page SEO, where most of the things still happen.

This post has become long enough, so let me split my overall content to another post to complete what I wanted to say and remove the gray uncertainties of SEO from many of your minds, particularly newbies.

[FredJones]

Now, let me head over to off page SEO. For Google, this is mostly about backlinks.

The better "incoming backlink juice" you get, the better off you are in off-page SEO. Incoming links cast votes for your site.

"All animals are equal, but some animals are more equal than the others" - George Orwell (Animal Farm).

But, this is not the case with backlinks. Not all votes have the same weightage. VIPs have heavyweight votes.

Who then are the VIPs in providing you backlinks? Pages that have high pageranks? Pages that are relevant? Pages that are highly viewed (high-traffic pages)?

First of all, number of views AFAIK does not matter in terms of leading you to good search results. Getting a backlink from a high-traffic page may mean high traffic for you though, do not undermine it - all I am saying is that it is not to be considered in the SEO context. In fact, for money-making, PPC is nothing but a form of placing backlinks on pages with good traffic and good relevance (to get laser-focused traffic), and then selling to that traffic. All I am saying that, this beast is completely different from SEO and deserves to be discussed in a different context altogether.

So, for SEO, I would consider two types of backlinks no matter there they come from (articles, directories, social bookmarks, profile links, blogs, home pages of other sites, whatever).

In fact, the overall ranking is determined by the on page SEO score, multiplied by the off page SEO score. And the off page SEO score is the result of multiplication of context in the offpage sense and page rank.

Context in the offpage sense is obtained by the incoming backlinks. The anchor text (or the alt tags of the incoming image links) and the relevance of the content of the page providing the link - both matters. I suspect that there is a combination of local context around the link being provided and global context of the page providing the link that actually matters (the local context around the link combined with the global "sense" of the article explains why even somewhat non-contextual links work very well, such as a link coming in your page about water while discussing about a desert).

Lets clear a huge confusion here. Does pagerank matter or not? Many people say yes, and many people say no. The no-school also refers to a video made by Google's Matt Cutts also, which appears to say that the page is not important/considered (people make their own version of understanding the difference between "important" and "consider" in this context).

Page rank matters. But, not unconditionally. Page rank (PR) matters only in combination with two major factors, which are not obvious from purely seearching a term and seeing search results (and that is why, just by seeing a page with a lower page rank appear over a page with a higher page rank makes many people feel that page rank does not matter).

In other words, if you see a search result produce a lower PR page appear above a higher PR page given a search term, and see that happening for many search terms, and end up deciding that PR does not matter, nothing could be more flawed.

Absolute value of PR may not always matter. But context of PR does matter. In other words, it is not the PR of the incoming link that is considered in isolation. Given a backlink, consider (a) the PR of the imcoming link multiplied (b) the context (anchor score) multiplied by (c) the PR of your page for that context - all three together. Sum these three factors across all links, and every single search result will seem to make sense, given the same level of on-page SEO. This is the driving science behind it. And there is nothing more to offpage SEO at this high level, and following these standards will get you to first page for 99% of the terms and following them really well will take you to #1 in most cases, given a good on-page SEO.

Thus, it is these 3 fundamental off page SEO factors and not one that determines whether PR is important.

In other words, there is no absolute answer to whether PR is important or not. It is important but only contextually.

When I see people arguing about whether PR is important or not, and both sides have their views, often they do it semi-blindly without realizing that they are talking of one factor while the actual factor is a deep embedding of 3 fundamental factors, one of which is PR. There are other subtle factors, but don't want to make it sound greek to most of you. This is a marketing forum and not a Computer Science forum, and I appreciate that.

So does PR matter? Of course, clearly. But not the absolute number (PR value) that you see on your Google toolbar. It is more than that.

The link context matters as I said above, and I shall work on PR only if it means the PR improves exploiting the link context, else it is utter nonsense (which, going by the locality and globalness of the context, does not have to be on a purely relevant page to be effective) and the anchor is the most important. The theory of LSI that had appliedd to on-page SEO applies equally well on anchor text of the backlinks. So the backlinks should majorly be on the main keyword but getting some LSI keyword links would work like magic. I have seen this work really really well, and have seen this work with some of my most knowledgable competitors also - my hats off to their knowledge (I figured this out only after I learnt my own way to look at every detail of every backlink of every competitor).

So, you should work on maximizing the combination of 3 factors, per link, mentioned earlier. And, remember that it is a sum-of-product (the well-known PI of SIGMA method, if you understand series mathematics), so for each link, all the three terms should score highly for the link to become really valuable. Anything shorter significantly reduces the value of the backlink.

Summary

Link context and good anchor ==> Good context score
High PR incoming backlink ==> Good contextual PR score
High PR of your page ==> Good confidence score

Multiplication of the three - a good score per incoming link.

Addition of all three across all incoming links ==> Your overall search engine result.

Actionables

Should you consider improving your PR?

I would consider that, but on a second priority. I would want to have sufficient contextual backlinks (backlinks with the right anchor text, and possibly from a page with related content to some extent though that is not mandatory) to my page. I would love to get a good PR, but not at the expense of the right context.

The other angle to this question would be, should you consider getting backlinks form high-PR non-relevant pages or low-PR relevant pages, given an option? Getting high-PR relevant backlinks is a no-brainer, so I am considering that you already take that you need them. By the way, are we not already saying that PR matters just by saying that we prefer high-PR incoming backlinks compared to low-PR incoming backlinks if the contextual relevance of the page providing the backlink is the same? Yes, we are.

Again, I would go for a mix of both. I have seen wonderful results by getting high-PR non-contextual backlinks, and a little lesser by getting low-PR contextual backlinks, but medium-PR comtextual backlinks also work wonderfully well.

The above has some tremendous real-life analogies. Let me give one hoping that it helps.

Who identifies the best soldier in the country?

If the defense minister does it, he would know the best. That is your PR-10 relevant in-context incoming backlink.

If the defense minister does not want to do it for you (you can not get a backlink from a PR-10 relevant link), what do you do?

Go to a highly ranked Major or Colonel and ask for this (get a meium PR contextual relevant backlink). Or maybe go to another minister and ask for it (get a high-PR backlink but from a non-contextual page). Both are good options. Both work almost comparably.

If none of the above such as a Colonel, a Major or another minister is available, what do you do? Climb down the chain till you get some good enough person (decent PR backlink) - probably you trust a local mayor (low-medium PR unrelated backlink) more than a local common-person (low-PR unrelated backlink). By the way, if you do know a common person with interest in defense, you would trust the person's knowledge as much as your mayor (low-PR but highly contextual backlink).

The above should answer the ever-important question - where should you get backlinks from given a chance - directly drawn from my experience. If someone had taught me outright I would have done faster, anyway thanks to my cool academic background I could make it out without too much effort and now I rank #1 in many terms.

I hope I could help all of you reading.

PS: Nowadays I take a shortcut - I watch what my competitors are doing both in terms of on-page and off-page - takes me a couple of hours to dig deep enough - and just emulate them for that niche. And then apply a little bit extra, which almost makes sure always that I come to #1.

[inter123]

Hi Fred.

Thanks for the info. I got a bit lost. Are you saying its a good idea to have lots of related terms to the keyword to get the 'TF' score up? And at what point is the cut off before becoming 'keyword stuffed'?

Originally Posted by FredJones

Term Frequency (TF) is a different beast. This is the what we call the "keyword density" - how many times does this appear on the page. Now, every LSI word has a similarity score with the main term, so every LSI occurrence improves the TF count without touching the term. The way I think Google computes keyword stuffing is by looking at TF (sufficient occurences but within a higher bound - and the longer the article/content is probably the lower the TF should be in terms of percentage), the score gained by the term due to LSI relevance (frequency of each LSI applied x the relevance score of that LSI), and adding this TF with the LSI-related score to come up with the TF-score.

[FredJones]

Originally Posted by inter123

Hi Fred.

Thanks for the info. I got a bit lost. Are you saying its a good idea to have lots of related terms to the keyword to get the 'TF' score up? And at what point is the cut off before becoming 'keyword stuffed'?

Sorry to have confused you - I am trying to capture every technical depth of SEO that I understand so it did become a bit too crammed with info.

I am saying, it is good to have a realistic mix of the main keyword and the LSI keyword.

What if you use too little of your keyword?

Simple - Google does not even undderstand that this keyword is what you are trying to talk about. So, it does not give you a good on-page score for that term and that hurts your ranking (see the second post in this thread to figure how).

What if you stuff your keyword?

Ideally, it would be a great relevance. And that is why keyword stuffing used to work earlier.

Then, search engines figured out that they have a great theoretical model implemented which is practically a cakewalk to manipulate (just stuff your keywords). Remember the older days when keyword stuffing used to work?

So they injected an artificial limit of keyword density (elastic to an extent) so that keyword stuffing is penalized. Industry researth makes believe that the limit is around 2% with a +/- of 0.5%-odd. This applies to your main keyword. I personally go for 1.2%-odd.

Because, what I would do it if I am writing lets say a 500-word article, I would use my main keyword around 5-6 times. But I would figure out 5-6 related keywords (LSI) and apply them emulating natural text. That would help the TF score rise overall by adding the main keywords TF (since it is inside the artificially injected permissible limit) and the LSI score that provides a leg-up to the main keyword (a fraction of the juice that the LSI keyword would have if that was the main keyword, practically gets sent to the main keyword and that helps the rank - the sentence became a little complicated but this is the simplest way I could explain it).

That would give a great on page SEO score.

Does that make it better?

[reynaerde]

Wow, VERY good read! I'm just starting out in IM but it feels like things are falling into place lately and this story certainly helped.

One question for you if you would be willing to answer:
For my first couple of websites (1+ month old now) I chose keywords (products, something like 'citrus juicer' say) where the main competitors have hardly any backlinks to the page (0 to 5 say), but they are ranking (I presume) purely based on content and authority (and internal backlinks?).
So the top 10 might have Amazon, about.com, target.com or other vendors in it (all subpages, not homepage, so amazon.com/category/product_bla_bla etc).
How would you go about ranking for such a term (no spying on backlinks here, there are none)?
Is it simple a matter of getting enough (sensible) backlinks or do you take another route to top those authority sites?

Thanks a lot!

[Davioli]

One of the few posts I've read on SEO that make a ton of sense. the way you've dissected the PR thing is almost perfect!.
Kudos mate.

[seoforu]

very nice post...thanks.

[Davioli]

Good Sig link plug ..

Originally Posted by seoforu

very nice post...thanks.

[FredJones]

Originally Posted by Davioli

One of the few posts I've read on SEO that make a ton of sense. the way you've dissected the PR thing is almost perfect!.
Kudos mate.

Thanks mate. In fact, I sometimes feel wild that these facts are so obviously simple but typically made to sound to over-complicated at times. I mean, I have literally read complete courses on SEO, including paid ones, that tell you things with far lesser depth. In fact, whatever paid and unpaid material of SEO I have read almost never goes into this much depth, which always surprises me - either the writers do not understand or they hide facts or they underestimate the reader. Each of the three cases are bad enough. So I spent more than an hour in writing the above couple of posts!

[bgmacaw]

I'll have to disagree with you somewhat on these on-page SEO factors...

Meta-tags
Meta-description
H1, H2 etc tags
Boldened, italicized and underlined content

I've not seen any of them make a difference in ranking for particular keywords, at least not within the past 2 years or so.

Meta-Tags seem to be totally ignored by Google, probably because they were so abused in the early days. However, Adsense seems to use them sometimes when the content on the page isn't text, such a in a picture or video blog.

Meta-Description does get used sometimes in the Google results so it's important for that aspect, but if it's missing Google doesn't seem to miss a beat.

H? tags and font styling don't seem to have a noticeable impact. I see far too many sites that are ranking #1 for highly competitive keywords that don't do the tags or font styles in 'SEO Friendly' ways. While it could have a minor effect, it seems that this is one of those "people aren't wearing enough hats" type distractions from what's really important.

On Outbound (relevant) links, I suspect this has been changing recently so that it doesn't count as much as it once did. It's obvious that throwing a few authority sites in your sidebar and/or footer without any surrounding text won't be helpful. Contextual links to related authority sites used to be a good idea but I've seen some things lately that make me wonder if this technique is being discounted now.

Oh, and you forgot to mention the new "holy grail" of on-site SEO, the site load time.

[FredJones]

Respectfully agree with many of the things that you said. But in most cases, they are different from what I am saying, and what you say does not conflict with what I say.

So, if I take meta-tags as an example, all I am saying here is that assume pages P1 and P2 have exactly the same on-page factors, except that the meta-tags are more relevant in page P1 compared to those in page P2.

In that case, the on-page SEO score S1 for P1 and on-page SEO score S2 for P2 will be such that:

S1 > S2

That's all I said.

What you are saying is that

|S1 - S2| < E,

where E is a small number. And I never disagree to that part. All that I am saying is that

(S1 - S2) > 0

Please take both the "| |" (mod value) and the "( )" (parenthesis) symbols mathematically along with other symbols, because I mean that way.

I completely agree that the value of E in the above equation has been reduced to a reasonably small number by the search enginew nowadays, and that is exactly what you are saying too.

In other words, there is no disagreement between what you and I say - we are saying two different things, that's all.

The same applies for say meta-description and many of the other on-page factors that you mentioned. But please note that, the fact that (S1 - S2) > 0 is still important enough when we are targeting that coveted #1 ranking and every bit is precious (given every other on-page and off-page factor to be the same).

About the quick loading factor - well, it does not come from the search engine theory. Practically I am still not sure - I have a huge site still hosted on a free server and I do still get the same rankings as I used to get last year (great ones, of course, including a couple of #2 rankings for a couple of related keyword for the same page, because a GoArticle holds the #1 firmly for both those keywords - and that GoArticle was written by Yours Truly lolzzz).

The moment I get a conclusive proof that the speed matters, I shall update - but without any proof did not want to confuse anything.

Great post anyway mate from you.

Originally Posted by bgmacaw

I'll have to disagree with you somewhat on these on-page SEO factors...

Meta-tags
Meta-description
H1, H2 etc tags
Boldened, italicized and underlined content

I've not seen any of them make a difference in ranking for particular keywords, at least not within the past 2 years or so.

Meta-Tags seem to be totally ignored by Google, probably because they were so abused in the early days. However, Adsense seems to use them sometimes when the content on the page isn't text, such a in a picture or video blog.

Meta-Description does get used sometimes in the Google results so it's important for that aspect, but if it's missing Google doesn't seem to miss a beat.

H? tags and font styling don't seem to have a noticeable impact. I see far too many sites that are ranking #1 for highly competitive keywords that don't do the tags or font styles in 'SEO Friendly' ways. While it could have a minor effect, it seems that this is one of those "people aren't wearing enough hats" type distractions from what's really important.

On Outbound (relevant) links, I suspect this has been changing recently so that it doesn't count as much as it once did. It's obvious that throwing a few authority sites in your sidebar and/or footer without any surrounding text won't be helpful. Contextual links to related authority sites used to be a good idea but I've seen some things lately that make me wonder if this technique is being discounted now.

Oh, and you forgot to mention the new "holy grail" of on-site SEO, the site load time.

[bgmacaw]

Originally Posted by FredJones

I completely agree that the value of E in the above equation has been reduced to a reasonably small number by the search enginew nowadays, and that is exactly what you are saying too.

Unfortunately, I see a tendency around here and some other discussion boards to place far too much weight on these page design factors (also W3C/XHTML compliance) that don't matter much, if at all, and not enough on things like links and actual content that do actually matter.

Originally Posted by FredJones

The moment I get a conclusive proof that the speed matters, I shall update - but without any proof did not want to confuse anything.

I know. Load speed is another thing people are worrying themselves about unnecessarily unless they have a real pig of a site. I figured I'd pitch that one out there for you to swat down since it was going to come up anyway.

[FredJones]

Again, I think you are talking complete sense here. I absolutely agree.

In other words, if I had to choose between optimizing between my content (or incoming links for that matter) versus my meta-tags and meta-description, I would any day go for content (or imcoming links) rather than my meta-tags and meta-description like things.

Practically speaking though, if I am designing my page, it is one of the easiest things to insert the right meta-tags and meta-description unless you are seriously pressed for time - it should not take anything more than say 5 minutes for most of the pages. So in general, one might as well spend those 5 minutes.

Not related to SEO, but I think meta-description is very important once you rank on the first page - since that gets shown to the person doing the search, and plays a tremendously important role when it comes to attracting traffic (but as I said, this is something completely different from SEO).

Originally Posted by bgmacaw

Unfortunately, I see a tendency around here and some other discussion boards to place far too much weight on these page design factors (also W3C/XHTML compliance) that don't matter much, if at all, and not enough on things like links and actual content that do actually matter.



I know. Load speed is another thing people are worrying themselves about unnecessarily unless they have a real pig of a site. I figured I'd pitch that one out there for you to swat down since it was going to come up anyway.

[FredJones]

I would suppose so. Build some good strong incoming backlinks, and do a good on-page SEO in the sense that have nice relevant content and some LSI keywords sprayed around. I would go with 1.2 to 1.5 LSI keywords per main keyword, and around 1.2% main keyword density. In other words, if you write a 500-word content, then I would say use the main keyword around 6 times and use 7-9 LSI in total (say, 4 LSI keywords, 2 times each, or maybe 5 LSI keywords, some once and some twice, such that the sum is around 8 times or so).

In your case, if you are certain that there is practically no backlink from the search result leaders but the leaders are established sites such as Amazon, I would go and do 10-15 social bookmarking in the top sites, maybe build 30-40 good incoming links (PR 4-6 or so), and syndicate around 8-10 articles over the next 10-odd days in the top 10- article directories. Should bring you at page 1 or pretty close.

Originally Posted by reynaerde

Wow, VERY good read! I'm just starting out in IM but it feels like things are falling into place lately and this story certainly helped.

One question for you if you would be willing to answer:
For my first couple of websites (1+ month old now) I chose keywords (products, something like 'citrus juicer' say) where the main competitors have hardly any backlinks to the page (0 to 5 say), but they are ranking (I presume) purely based on content and authority (and internal backlinks?).
So the top 10 might have Amazon, about.com, target.com or other vendors in it (all subpages, not homepage, so amazon.com/category/product_bla_bla etc).
How would you go about ranking for such a term (no spying on backlinks here, there are none)?
Is it simple a matter of getting enough (sensible) backlinks or do you take another route to top those authority sites?

Thanks a lot!

[reynaerde]

Thanks for your answer, encouraging to hear that I'm on the right track!
Just one clarification: by '30-40 good incoming links (PR 4-6 or so)', do you mean actual page PR 4-6? Or domain/site PR? Would blog comments count or are you talking about in context/text links here?

Thanks again!

[Mike Anthony]

Fred I am in two minds in regard to this thread. First I think you did a good job in the over view. The balance of on page and off page SEO, the introduction to LSI and keyword placement - all good. Most of all of that has been coverd here multiple times but one more time is never bad.

Heres the bad. Theres rampant speculation stated as fact (and claiming to be a guru doesn't help). If you are as conversant as you claim then you must know that there is considerable disagreement in the SEO community about the relevance of the referring page. There is little definitive evidence for it. I lean in that direction but mark of a good professional is not overstating our leanings as facts.

Second there isn't much controversy AT ALL that Meta tags are of VERY little value. Thats widely known and been confirmed on mutliple occassions by Google so that part of your post was strange. A little less strange was the Bolded comment. As you should be aware often times (not always and I wouldn't even say most) boldness is specified in CSS and I 'd like to see the evidence that Google in particular concerns itself with what essentially is the style of the content. You are going to have to do better in proving that or its just a merry dance.

Then you get to the mathematics. As a programmer I know how we like to imagine how things work and apply how we would do this or that. Its in the blood but we have to be careful. Neither you or I have access to any search engines algorithm. We CANNOT claim that our knowledge of computer science translates into an understanding of an unknown algorithm. That would be nonsense and trying to fool the uninformed. There is no magic search engine theory that opens the door on Google , Yahoo or MSN. Gives us hints yes. Give us particular scores - Balderdash,

A programmer of a search engine can include all kinds of things in their algorithm that you can't see in the output. Claiming to know the total score and how to calculate it is just emperor with no clothes stuff.

So in general good stuff but you take away from the value of the good stuff by launching into conjecture and stating it as fact.

What in particular am I calling into question? A quick list of the ones that I think will end up in newbies wasting a lot of time.

A) That you can duplicate the ranking of a page on an article directory by doing the same on page seo. Article directories do not rank because they have superior on page seo than everyone else. Pages on them rank because they have high authority and have mega backlinks.

B) That there is a context PR distinct from general PR used in any calculation (I'd like to see the data on that) and that that equation can get you to the top spot 99 percent of the time.

C) ( related to B). Whenever I see any SEO claim that they can get number one "almost any time" in any niche I know beyond a shadow of a doubt that he/she is in full guru selling mode not reality. This is really dangerous to a newbie. They pour their heart and soul into your algorithm busting calculations and then burn out when it can't deliver. NO one on this board knows all the factors that go into a rank and therefore no one can honestly claim to have an almost guaranteed way of getting to number one.

I'll give you some mathematics to illustrate that point

10 + (T X 2) + 11 + (Y X W) + 14 +1 +98 + 12 + P = 21,200

define W.

Can't? Why? Because the unknown variables are too many to make a definitive statement. You cannot claim to know how to rank number one almost every time in any niche as if you know all the factors. Its a misleading overstatement no professional SEO would make. furthermore you cannot even duplicate your competitions rank by duplicating what they did. They can have links on sites that you can't get. have domain age, aged links and all kinds of variables.

I'm sorry but some of your claims just come across to someone else with a computer science and SEO background as sales talk for your services. I would never claim that my background gives me xray vision into Google offices where their algorithm is spelled out.

[FredJones]

Mike,

Great post and good points.

I really don't claim any knowledge into Google's or any other search engine's backoffice - because I don't have that. And I do realize that you have god points when you say in your example, "compute W" from a single equation with multiple variables. I am not trying to do that anywhere.

In most of the cases, I am claiming that given that all your other X's and Y's are the same, here is your U and Z and so on that you would want to increase, and here are the impacts of increasing them separately versus simultaneously. An example of this is my emphasis on the PI of SIGMA (sum of product).

Seriously speaking, can you site an example where you have seen what I said there not work? Also, could you please explain the individual parameters, and prove that the pages outranking your example for the chosen keyword do score lesser in the Sum-Of-Product that I have mentioned? I would really love that, accept that as the only direct proof against what I said on off-page SEO (unless you can point me to other ways to contradict my understanding, which I shall happily accept if you reason it out well enough) and would be a great lesson for me.

I never claimed that things such as domain age are important or that they are not important. This is because, these factors are too heavily dependent on individual search engines and not baed upon the fundamental principle of search results (in my understanding), though it is said by many that search engines use domain age and some other parameters. Personally I have no clue whether it matters or not.

As per the in-context PR is concerned - I am sure you would appreciate that if the context as determined by the anchor text did not matter, any page with PR X would always rank above any site with PR < X given a search keyword K (under the ideal world assumption that the on page SEO score for K is the same for both). There are too many exceptions to this.

Taking Google as an example, and in particular the PR factor being totally attributed to Google, I have gone into depths of what search engines look for, in my understanding. I do realize that you have every right to disagree, and I am sure you have your own justifications for that just like I have mine and I totally respect that, but I was just trying to bring out a combination of what has worked extremely well for me, how I have broken that up and how well that matches the fundamental understanding of Computer Science that I have. That's all.

I politely request all the newbies to distrust my post if you want to since what I am sharing is my understanding (on which I have deep enough personal faith) and do not guarantee anything for you. Please proceed by your own understandings. I have just shared what I said above - a combination of (a) what works for me, (b) how I break that up and (c) how that matches with my fundamental understanding of the process from my Computer Science educational background, which, in my understanding, is excellent (and now please let me tell further something that I did not want to hype earlier, I have taught Computer Science to many brilliant students in India, including someone who was a President's Gold Medalist in his B. Tech in India and had gone on to become an outstanding performer during his PhD in Computer Science in Stanford University, and other people who went on to do PhDs in Computer Science from the top-10 universities in the world). But all these tall claims do not mean anything in terms of of how actual search engine algorithms are implemented because I have personally never seen any, so please implement at your own risk, only if you think you want to. At the same time, please note that I am indeed trying to give my best here.

By the way, Mike, I beg to differ about the article directories thing that you said on rankings. I do not discount the links, but again, all I am claiming there is that (ideally) if you take an article directory D and a website S with same PR and exactly the same off-pave and on-page SEO score, and then create a page in S with the same content as what you write in a new article that you post in D, then the article in the article directory would practically be seen to outrank the new inner page in reality, and the reason behind that would be that typically the article directory would have done a better on-page SEO by virtue of its framework. But, if the website S has the same quality of on-page SEO then it would go to other tie-breaking factors (I don't know what they are - here is where my and your and everyone else's lack of knowledge into the search engine backoffices kick in - but this may be factors like domain age, page load time etc - I repeat I don't know what) to rank one above the other. In practice, such a close tie has an extremely rare probability and I'm sure you would appreciate that.

I felt shaky when you said the meta-tag has low value - did you read through the entire thread? It is not too long yet I hope! Not the original couple of posts but the following portion of the thread. In response to a subsequent post, I have explicitly said that the value has to be positive, with a quantity that I do not know how much. Again, if you go back to my two original posts after reading the subsequent post on meta-tags, you would realize I am never contradicting the statement. When you said "VERY little value" is that not vague anyway? My statement that it is > 0 is much more concrete than that - you statement does not even tell me that the "VERY little value" is a positive value - does it? It is a positive value in my opinion, of a possibly small quantity.

Whenever I see any SEO claim that they can get number one "almost any time" in any niche I know beyond a shadow of a doubt that he/she is in full guru selling mode not reality. This is really dangerous to a newbie. They pour their heart and soul into your algorithm busting calculations and then burn out when it can't deliver. NO one on this board knows all the factors that go into a rank and therefore no one can honestly claim to have an almost guaranteed way of getting to number one.

About the selling mode that Mike talks about - lets just say that I am not trying to sell. Anyone reading this thread, my sincere request, please do not buy anything from me anywhere, on Warrior Forum or Outside, because of this thread, be it something related to SEO or be it anything else. Please treat this thread as an informational one only. Please trust/distrust this information wisely, and please act only at your own responsibility. I am just laying out my own opinions on SEO here, and it need not be perfect (though I have given the best that I know).

By the way Mike, about the fact that many things have been covered before that I mention - true - but could you or anyone else reading this please show whether the sum-of-product factor has been covered in the sense that I mentioned there? I would love to read related literature, that I have not read.

And please note that I know very well that PR is computed as a recursive sum of product, which is completely different from the sum-of-product that I mentioned above - so please don't refer to how literature talking about how to compute pagerank in response to this unless you think you are referring to something special. This literature is very much available in a number of websites, and other text documents.

Finally, thanks once more for your opinions and great post (I shall press the Thanks button once I finish posting this), and I seriously value your opinion which is why I spent more than 50 minutes writing this reply.

EDIT: By the way, the context of the backlink matters, and a great example is "click here" for Adobe - I believe they would rank at #1 with far lesser "equivalent distribution of backlinks" (if you know what I mean, because I know I don't knnow the right way to express this in English) for the term "Adobe" or maybe "Adobe Reader" than "click here".

Also, a general observation - many of the readers of this tread are trying to infer my opinions about absolute values of importance of attributes (such as meta-tags) where in the first place I did not express any opinion for many attributes but did express my opinion on some attributes in the first two original posts andd pretty much the subsequence posts also. Please do not try to read between my lines unless you are certain you know what you are reading, you may not read the right message at all.

[FredJones]

Mike, now that I look back at the math example that you used, let me say that you did not formulate the same problem that I did. I did that from a model building perspective and you did that from a problem solving perspective. What did II mean by that? I mean the following.

10 + (T X 2) + 11 + (Y X W) + 14 +1 +98 + 12 + P = 21,200

Can I compute the value of W? No. By the way, you said define W, and I am taking that to be "compute the value of W" - because I do not understand how to "define W" from this - you do not define a variable from an equation - you compute the value of the variable from an equation.

Assuming that, I move ahead.

10 + (T X 2) + 11 + (Y X W) + 14 +1 +98 + 12 + P = 21,200

=> W = (21000 - 10 - (T x 2) - 11 - 14 - 1 - 98 - 12 - P) / Y.

So, the spirit of my post is that, if maximizing W is your target (like maximizing your SEO score), then you should try to:

-- Minimize Y
-- Minimize P
-- Minimize T

I never say how much (quantitatively) to tweak Y, P and T, because that is where the search engines have their individual sauces, but assuming rational behavior, I would want to say minimize Y, T and P and so on. That's all I was trying to say in my post (in those lines, in some sense but not an exact sense as an optimization problem).

Then I had added some other things that I believe have been proved well enough to satisfy me, such as the maximum keyword density which is an artificial limit imposed by search engines for real-life protection of obvious and easy manipulation of search rankings.

That's all. One would hurt their own legs if they try to read anywhere more than that. I meant exactly what I said there. I may have missed things (recall may not be 100%) but I hope the precision of what I said if you take that literally is 100% or close to it (just my hope, please disagree if you want to and I shall totally respect that).

[Mike Anthony]

Originally Posted by FredJones

Mike,

In most of the cases, I am claiming that given that all your other X's and Y's are the same, here is your U and Z and so on that you would want to increase, and here are the impacts of increasing them separately versus simultaneously.

the x and ys are never the same thats the entire point. G can affect U And Z and on and on. When we don't know something then we ought to just come out and say we don't know it. Thats the issue I have with newbies reading your almost always rank #1 claim. Its just false.

Also, could you please explain the individual parameters, and prove that the pages outranking your example for the chosen keyword do score lesser in the Sum-Of-Product that I have mentioned? I would really love that, accept that as the only direct proof against what I said on off-page SEO

Certainly we can get to that if it warrants it but first you have to present some proof FOR what you are saying then I can take a look at it and we can go from there. The nature of forums is that anyone can come on claim to be anything and make any kind of assertion. I specifically asked you your evidence for Context PR separate from Normal PR displayed in say Google's toolbar (as an example). Its a key part of your equation and if you can't confirm it is separate from toolbar PR then that entire paragraph falls apart.

I never claimed that things such as domain age are important or that they are not important. This is because, these factors are too heavily dependent on individual search engines and not in the fundamental principle of search results.

I'm sorry but thats meaningless. No one ranks their site in the fundamental principle of search engines. They rank their sites in Yahoo, Google and Bing so of course you have to look at the individual search engines. You can't come up with anything that works without looking at the individual search engines and given the present SE market - Google in particular. Its pointless otherwise. Incidentally yes Domain age and even backlink age is important . Those are two pretty well known metrics that have been tried in the field and confirmed in the case of Google by looking at their historical data patents. You need to study these things because not knowing them can lead you to come to false conclusions about the factors you are looking at.

As per the in-context PR is concerned - I am sure you would appreciate that if the context as well as the anchor text did you matter, any page with PR X would always rank above any site with PR < X given a search keyword K (under the ideal world assumption that the no page SEO score for K is the same for both). There are too many exceptions to this.

No it would not because

A) the content on the page would make the difference and the anchor text coming in would factor. That is not Context PR. Thats just content. NO need to make up another kind of PR score. Its totally unnecessary.

B) As Google has confirmed PR is not the end all and be all of ranking. Theres is some evidence that Google has a factor of authority that though tied to PR is not the same a PR. That ties in nicely with what many have seen here with profile backlinks - that the domain can have authority even when the page is near zero PR.

, I have taught Computer Science to many brilliant students in India, including someone who was a President's Gold Medalist ........... But all these tall claims do not mean anything in terms of of how actual search engine algorithms are implemented

Just one question. IF the tall claims are not important then why bring them up in an entire long paragraph? Thats an incoherent conclusion after a long paragraph of listing them. Thats why I siad and say you are in full guru mode. I'd like to really make this thread useful by having you put some fact on the table. Lets looka t some real life serps - but so far as I see the only facts you are even trying to put on the table are your alleged credentials.

Where are the serps where a site ranked high as ezinearticle by doing the same on page SEO? You cited that as a key reason article sites rank higher than other sites AND you stated on the basis of that that anyone can rank just like that (if you want to pull back on that thats fine. I can appreciate the honesty). Where is there the evidence of a context PR that is separate from normal PR? How can you claim to know what the total score is when you don't know the algorithm?

Like I said some good stuff but the down side is also a whole lot of conjecture as fact without putting any real facts on the table. I'm asking you to nail down this stuff precisely because you claimed this was REQUIRED reading for newbies and so far as I can tell some for them already in jeopardy of wasting time on a very incomplete model of unserstanding real world search engines like Google.

[FredJones]

Lets not get into arguments, but I do appreciate that you have some excellent points here. Trust me, I am busy enough in my own life and I am sure that you are, so with every due respect, let me summarize this with 3 things.

1. I do get to #1 using this and inside me I believe others would - but I appreciate what you are saying and so I have edited my original post to request anyone following this use their own sense rather than blindly sticking to my words. Hope it helps avoiding confusion among others? By the way, no offense meant, but I do feel hurt when you say my claims on #1 is "just false" - wish it was said differently.

2. I would love to get a read of the historical data patent - then I can conclusively add that to this thread. Could you please point me to the literature, since you have already read it / read about it?

3. Let me change the title of the thread to make it different from the "must-read"? Here I believe this is my mistake - I was just trying to give what I thought to be the best, so said "must" but of course nothing is "must" in the world!

About my "alleged" credentials, I don't need to prove anything really on that. I was just speaking the truth - taking or leaving that is anyone's choice again, just like anything mentioned in my post.

Thanks!

[Mike Anthony]

Originally Posted by FredJones

That's all. One would hurt their own legs if they try to read anywhere more than that. I meant exactly what I said there. I may have missed things (recall may not be 100%) but I hope the precision of what I said if you take that literally is 100% or close to it (just my hope, please disagree if you want to and I shall totally respect that).

Fred I don't know how more precise I could be. I've said multiple time that my problem was with you stating almost always in regard to ranking number one. there is no such magic pill for SEO. You do not have the mysteries of the Search engines unravelled and the way that you wrote that before the edits was misleading. I'm going to stick with your #1 claims as false. If you get number one all the time using any technique its because you don't wade into the deep water of real competition. I don't know ANY Seo even the top names that know more in their pinky that you or I that get those reults in every niche. So until proven other wise I am going to call that bluff.


I don't know how much you teach seo to newbies. One thing VERY common and destructive thing in IM is the amount of people that get setup for false expectations. they figure they have the magic secret now so they are assured of overcoming any serps and they WASTE lots of money and time climbing a pointlessly steep hill because they believe they have the magic potion.

For example there was one poster in this thread that listed being new and up against amazon, Target, about.com kind of sites etc. After reading your posts he concludes he is on the right track (after all since you almost always rank number one for any serp why shouldn't he?). Could he be if he gets these high Pr links you suggested (I'm sure he doesn't know how based on his followup question)?

Possible. My bet? He's chosen the wrong niche (can't say for sure since we don't know the details). A brand new site is going to struggle against those kind of sites. The have been established in the serps for years and have aged domains and authority. He will more than likely spin his wheels. money and time. BUt he'll do it because he figures he has the magic potion - only he doesn't.

Nufff said. theres Sunday dinner and my team is in the NBA conference finals. enjoy the rest of your weekend and thank you for the candor in your edits.

[actionplanbiz]

Thanks FredJones for the Info. but what do you got about getting backlinks from .EDU and .GOV? does it need to be relevent? high PR?

[Mike Anthony]

1. I do get to #1 using this and inside me I believe others would - but I appreciate what you are saying and so I have edited my original post to request anyone following this use their own sense rather than blindly sticking to my words. Hope it helps avoiding confusion among others? By the way, no offense meant, but I do feel hurt when you say my claims on #1 is "just false" - wish it was said differently.

You know what enough of this posturing. Tomorrow we can put up a test. Pick a highly competitve term and lets see it. Since your saying you believe anyone can do it (and therefore across all niches) then I'll pick a highly competitive term and when you rank number one then the discussion will be settled.

edit: Never mind. I just looked at what the strategy is that you are selling. Its not new and it doesn't work for most serps

[FredJones]

Yes, a high-PR .edu or .gov is better than low PR .eddu or .gov. And by the way, I mean the PR of the page and not the domain.

Going back to relevance, please look at factor "(b)" that I have mentioned in the off-page scoring. The page does not need to be "relevant" in the sense that I see people talking about, though having a relevant page can not hurt and I believe could set off some sort of a global relevance score.

But the local relevance, which means the anchor text of the hyperlink, must match. Which means, if you are promoting a page related to toys, it is good to have toys or anything related to toys in the LSI sense to the hyperlink rather than mentioning the anchor to be books for example.

Addditionally, I suspect, but without proof, that wrapping the link in a sentence (or small paragraph) with a related note which has semantic relevance (determined by LSI/main keyword) would help.

Originally Posted by actionplanbiz

Thanks FredJones for the Info. but what do you got about getting backlinks from .EDU and .GOV? does it need to be relevent? high PR?

[FredJones]

Originally Posted by reynaerde

Thanks for your answer, encouraging to hear that I'm on the right track!
Just one clarification: by '30-40 good incoming links (PR 4-6 or so)', do you mean actual page PR 4-6? Or domain/site PR? Would blog comments count or are you talking about in context/text links here?

Thanks again!

I have seen text links to work better than blog comments in recent past. However, there is no theoretical background for this and either may be an exceptional case for me or may have something to do with how the search engines tweak their code to avoid artificial manipulation or whatever.

I meant actual PR 4-6. The domain having PR 4-6 is good because it keeps a chance of the PR being in-flown into the inner pages, but I love actual PR 4-6 over domain PR 4-6.

[FredJones]

Mike,

Good luck to your NBA team.

I have done my couple of changes and hoping that you shall provide the historical data patent - because hearing at the air level does not make sense.

Please maintain whatever opinion you want to about me - it is your independent wish.

But let me say this - I have no intention to compete with you or with any termr that you have chosen just to prove this. After your post, I have clearly edited to make sure that whoever reads it uses their own judgment, and all that I have said (does not make sense to repeat the whole thing).

edit: Never mind. I just looked at what the strategy is that you are selling. Its not new and it doesn't work for most serps

I repeatedly tell this, forget what I am selling elsewhere - this one is an informational thread. Please do not buy anything from me because of this thread (applies to anyone reading this thread), and please stop confusing between what I try to inform and what I sell. They are different. And please stop digressing away this thread's focus with all your sell phrases and your rants on that you don't know how much SEO I teach to newbies (so far I don't, but that is irrelevant) - I feel tired of it now.

If you have anything more productive to contribute (for example, what has worked for you on SEO, or a real reference to the historical data patent by Google), please do that, either here or elsewhere. I would appreciate that and knowing Warriors, almost all of them would also appreciate that.

[Mike Anthony]

Originally Posted by FredJones

your rants on that you don't know how much SEO I teach to newbies (so far I don't, but that is irrelevant) - I feel tired of it now.

Good then you have a better sense of how tired I was of your self proclamations and who you have taught. Its quite apparent that you have no intention of putting anything concrete on the table so we are both tired of it. We end on an agreement.

[FredJones]

Originally Posted by Mike Anthony

Good then you have a better sense of how tired I was of your self proclamations and who you have taught. Its quite apparent that you have no intention of putting anything concrete on the table so we are both tired of it. We end on an agreement.

Sure, we end on an agreement. I shall just keep saying what has worked for me, and my beliefs of why they worked, but I shall not give any screenshot and shall not invest any time in proving that my thing works (because my results anyway tell me that and hence I am happy to believe it), till anyone disproves it which will be a learning excercise for me also. So if that means not bringing anything more concrete to the table, then I would be happy not to bring anything more concrete on the table.

Also, wanted to clarify the taught - I have taught Computer Science (the core science) to B Tech (equivalent to BS in USA is B Tech in India), M Sc (Masters of Science) in Computer Applications and M Tech (Master of Technology, equivalent to MS in the USA) - a number of theory and practical subjects. SEO is clearly not a subject taught at anywhere that I have studied or taught, and I have not taught SEO to people ever.

By the way, hypertext processing was taught in the institute that I did my Masters by the famous and respected Soumen Chakrabarti (Soumen Chakrabarti), who happens to be a senior PC member even this year in SIGIR (2010) and was the Area Chair in SIGIR 2009. I did not take his course or work with him, though I had taken the machine learning and data mining courses as a student. This is only for clarification since it came down to SEO getting or not getting taught, nothing more.

[Mike Anthony]

Originally Posted by FredJones

By the way, hypertext processing was taught in the institute that I did my Masters by the famous and respected Soumen Chakrabarti (Soumen Chakrabarti), who happens to be a senior PC member even this year in SIGIR (2010) and was the Area Chair in SIGIR 2009. I did not take his course or work with him, though I had taken the machine learning and data mining courses as a student. This is only for clarification since it came down to SEO getting or not getting taught, nothing more.

negative . It did not and does not come down to getting or not getting taught. It came down to you not wanting to put any evidence on the table.

[Fernando Veloso]

Originally Posted by Mike Anthony

negative . It did not and does not come down to getting or not getting taught. It came down to you not wanting to put any evidence on the table.

Mike,

What evidences did you bring to this thread? References? Tests? Data? Didn't see any.

Just a discussion on how your opinion is different than his opinion.

You might want to take some time to realize people has different opinions based on their tests/skills/results but they are not obliged to show you or me what we always want: data we can track.

Did you ever tested the methods Fred is talking about?

[louisejessica5]

Good.......

Nice views on SEO beliefs

I have gained so much frm these

[jazbo]

I totally agree with Mike. This adds nothing new and has questionable advice. As for the completely hypothetical calculations - worthless and confusing.

And to link to this thread in your sig as an "SEO Knowledgebase" does not add to your low credibility.

SEO is simple. Write lots of relevant content and get some decent links to it. What peoploe do NOT do is STICK AT IT.

[Mike Anthony]

Originally Posted by Fernando Veloso

Mike,

What evidences did you bring to this thread? References? Tests? Data? Didn't see any.

Hey Fernando

As you are very aware I've bought plenty to this forum by way of references/data some of which has annoyed you in the past because it contradicts your opinions. In this thread I wasn't claiming my opinions "were required reading" that I could rank a site numbr 1 "almost every time". that my knowledge of programming or computer science translated to me knowing the equations that Google uses. etc etc.

Many tall claims while your are linking to your SEO WSO should come with some evidence

Incidentally you might consider practicing what you preach. Can you point to any thread where you have added any data? Any thread?

[FredJones]

Originally Posted by jazbo

I totally agree with Mike. This adds nothing new and has questionable advice. As for the completely hypothetical calculations - worthless and confusing.

Your call - take it or leave it - I am sharing what works for me and have advised people to work on their risk. I still believe this in my heart of course.

By the way, as I understand it, every single SEO suggestion that people give is based on what woks for them, since none have the real insight into the actual search engine algorithms (and if they do, then they are under legal NDAs so they do not speak). So, why is what works for me supposed to be all that worthless etc, while others are not? Or are you saying that every SEO suggestion in the world is all of that too? Goodness !!

And to link to this thread in your sig as an "SEO Knowledgebase" does not add to your low credibility.

Sorry? My low credibility? I think this observation is questionable, hypothetical, worthless and confusing. Just my personal opinion, again take it or leave it.

SEO is simple. Write lots of relevant content and get some decent links to it. What peoploe do NOT do is STICK AT IT.

Lots of relevant content is great, but some decent links may not be enough unless you know exactly what you want your link to do (which is not much difficult, I agree). But if you realize, I had mainly addressed SEO newbies and this may not be obvious to them.

[levanhen]

nice sumary SEO, I'm like "Content is King", It good for every time.

[reynaerde]

Just to add from the newbie perspective:
It is indeed quite confusing to hear some people proclaim that you should stay away from Amazon, target, etc top 10's, because of their authority where other people tell you the exact opposite being that usually the ranking pages on such sites are not optimized and won't have any backlinks pointing at them so (if that is indeed the case, 0-5 backlinks) it is really quite doable to get first page rankings.

Obviously I have never tried this out myself, so I have to rely on what others are saying and what makes sense to me (which is of course sculpted by what I read). So.. basically I won't know for sure until I try it I'm afraid.. Or I could set up a poll, haha, 'do you think this keyword is doable'. Hmm, actually not such a bad idea

[FredJones]

Thanks for sharing your perspective, it adds value.

Personally I have beaten Amazon a number of times. That does not mean much though - unless you know exactly how strong their page is.

Do you see Amazon rank #1 in all that they do? No! So they can be beaten.

How hard is it to beat them? Some pages are easy, some are hard. Depends on the exact page and niche that you are planning to attack, as well as the exact keyword. Analyze the page in-depth with respect to the keyword and go ahead if you gather enough confidence.

In other words, exactly how much oil it takes to burn a lamp can be described, but never felt until experienced - but stick to the basics of SEO and you would be well-rewarded is the general principle.

Originally Posted by reynaerde

Just to add from the newbie perspective:
It is indeed quite confusing to hear some people proclaim that you should stay away from Amazon, target, etc top 10's, because of their authority where other people tell you the exact opposite being that usually the ranking pages on such sites are not optimized and won't have any backlinks pointing at them so (if that is indeed the case, 0-5 backlinks) it is really quite doable to get first page rankings.

Obviously I have never tried this out myself, so I have to rely on what others are saying and what makes sense to me (which is of course sculpted by what I read). So.. basically I won't know for sure until I try it I'm afraid.. Or I could set up a poll, haha, 'do you think this keyword is doable'. Hmm, actually not such a bad idea

[Mike Anthony]

Originally Posted by reynaerde

Just to add from the newbie perspective:
It is indeed quite confusing to hear some people proclaim that you should stay away from Amazon, target, etc top 10's, because of their authority where other people tell you the exact opposite being that usually the ranking pages on such sites are not optimized and won't have any backlinks pointing

True it is confusing but thats why you should take things said on a forum with a huge grain of salt. Especially on an IM board where people tend to make money off of selling themselves as Gurus. No one said you can't beat Amazon or that Amazon ranks high on every term. Thats just a distraction. Fred isn't addressing himself to your question and example. Its not Amazon its the combination of all the other authority sites in your example. Its not that you can't rank its that if you do the research you can get better terms with quicker results

This is how people waste newbies times but yes sometimes it can pay off. Lets look at your example

"Citrus juicer"

Sure enough theres Amazon at the top with it seems very few links and another link from them under that (don't have the time to run it through all the checkers) then theres Google shopping results. Good luck bumping that off the first page. Then theres Cusinart one of the top names in juicers

So thats the top four After that is discount juicers and Cooking .com and target. Discount juicers has a lot more backlinks than at least one of the listing above it and my guess is the links are not all complete garbage since its a PR3. Amazon is showing as a Pr1 for that page. So can you do it?

Its possible but you will have to move over a PR3 page with decent backlinks on your one month old site and then you need to beat Googl'e own shopping results and then take out Amazon and also Cuisinart as PR4. They are all optimized for the keyword and before anyone starts on me about PR. no it doesn't rank by itself but it IS an indicator that can be used with the quantity of a site's backlinks to determine a preliminary quality of the links.

Possible success? Yes. Would I send a newbie to do that as his first Internet mission. Nope. Why? Well three things

A) I'm not entirely sure why the rankings are what they are without more analysis. Google doesn't tell me all the backlinks its sees and just a few really good ones that I DON"t KNOW about could lead to frustration. What happening with trust and authority here again don't know without diggin deep.

B) the popularty of the sites for this term would leave me uncertain of long term success. Cusinart has a sale and picks up organic backlinks from people referring to it with in content high pr pages (Fred seems to think you have highquality PR4 and up backlinks in your back pocket) and I may be out of my money slot in no time. Its not an obscure term its the key name for a popular product and theya re a known name for that product

C) I normally tell people that are new to go and make some money and build themselves up. Thats a pretty competitive terms with over 134,000 searches a month with people gunning for it. The effort you put into that you could rank for two or three long tail keywords and be making some money.

Now I kow thats only your example but thats all we have to work with. Your real keywords could be less or worse in terms of competition.

[SurviveUnemployment]

Hey Fred! Thanks for the interesting take on it.

[reynaerde]

Although there are different opinions vented in this thread, it still is very informative to me, you're basically both making sense to me so thanks!
Anyway, I will actually try both: try to rank a less competitive term and the one I was referring to (which is actually less competitive than the 'citrus juicer' example, couldn't find a closer match to my kw). Best case scenario it will just take me longer to rank for the more difficult term but I'm sure I'll learn a lot along the way nonetheless

[FredJones]

I did not and still do not see Citrus Juicer anywhere in your question - where did that term come up from? Did I miss it?

Also, your answer does clarify that probably Citrus Juicer was not your target anyway. Hope you rank well for what you do.

Forget relying solely on Google backlink checking. As Mike also correctly mentioned above, Google misses showing too many. There are a number of places that you ought to check backlinks from. Personally, I would go for Yahoo, Google, Altavista and AllTheWeb - all 4 together - at the very minimum. Individually do not trust Yahoo site explorer too - because it shows different numbers at different times (at least that is what I see, for the same site).

Also, backlinks are nothing if you do not understand the context of the backlinks, primarily set by the anchor text (please look at my second post of this thread). So for example (I know it is not your example), if you are trying to rank the same page both for "mini fridge" as well as "mini refrigerator", and if you see a single page already ranking for both the terms, then there comes your perfect guide who can show you exactly what it would take you to rank for both the terms using the same page (assuming that it is what you want, as an example). Analyze this page - their backlinks and the anchor texts they have. It will probably become clear to you on what mix of "mini fridge" and "mini refrigerator" on what kind of pages (pagerank etc) gave them this kind of rank.

See it? This is your clearest way to rank well and quickly for both the terms. I'm sure if you keep working without this analysis, you still may rank well for both terms and with luck quickly too, but then it will probably be more like an an unguided missile hoping to hit target since it was thrown Eastwards and your target was Eastwards too. Spend a bit of time studying, and purely by your own efforts you can figure out exactly what may make you come to the top 10 etc.

The above methodology is the result of the good old sum-of-product computation that I believe happens at the background for every single page I have ever come across. Most people never realize this (as expected) and some people here fail to appreciate after being pointed out (which is okay for me anyway).

Originally Posted by reynaerde

Although there are different opinions vented in this thread, it still is very informative to me, you're basically both making sense to me so thanks!
Anyway, I will actually try both: try to rank a less competitive term and the one I was referring to (which is actually less competitive than the 'citrus juicer' example, couldn't find a closer match to my kw). Best case scenario it will just take me longer to rank for the more difficult term but I'm sure I'll learn a lot along the way nonetheless

[Mike Anthony]

Originally Posted by FredJones

I did not and still do not see Citrus Juicer anywhere in your question - where did that term come up from? Did I miss it?

Also, your answer does clarify that probably Citrus Juicer was not your target anyway. Hope you rank well for what you do.

Post number five. right there in the second paragraph. OF course it wasn't his target but that was the example so thats all anyone can or at least should comment on.

The above methodology is the result of the good old sum-of-product computation that I believe happens at the background for every single page I have ever come across. Most people never realize this (as expected) and some people here fail to appreciate after being pointed out (which is okay for me anyway).

NO people who know SEO (and sorry apparently better than you do based on the things that you were unaware of) just called you out on the total bogusness of claiming to know what Google's algorithm was and is and they will continue to do that. Despite claiming otherwise you continue to push your unsubstantiated beliefs as superior facts that everyone else is unaware of or too uninformed to appreciate. It is HIGHLY deceptive and newbies need to be made aware that no amount of alleged computer science background gives either you or me credentials (even if they could be verified which they can't) to claim to know what the details of Google's secret algorithm is or to claim that supposedly knowing some alleged facts of yours allows you to rank number one for almost any serp you choose.

You've flat out refused to provide any evidence that you can do that and its therefore bogus to claim that there is anything superior about what you offer that other people cannot grasp. This continues to be posturing and very little else. Its nothing personal like I said it is just entirely too deceptive for me not to continue to comment on if you continue to make or imply those claims. You simply need to quit it. The only pointing out you are doing is to your self as the alleged expert and guru (while coincidentally selling your product on the same basis) not to even a single piece of independently verifiable evidence..

[FredJones]

Originally Posted by Mike Anthony

Post number five. right there in the second paragraph. OF course it wasn't his target but that was the example so thats all anyone can or at least should comment on.

Good - thanks for pointing out, I had missed it.

NO people who know SEO (and sorry apparently better than you do based on the things that you were unaware of) just called you out on the total bogusness of claiming to know what Google's algorithm was and is and they will continue to do that. Despite claiming otherwise you continue to push your unsubstantiated beliefs as superior facts that everyone else is unaware of or too uninformed to appreciate. It is HIGHLY deceptive and newbies need to be made aware that no amount of alleged computer science background gives either you or me credentials (even if they could be verified which they can't) to claim to know what the details of Google's secret algorithm is or to claim that supposedly knowing some alleged facts of yours allows you to rank number one for almost any serp you choose.

You've flat out refused to provide any evidence that you can do that and its therefore bogus to claim that there is anything superior about what you offer that other people cannot grasp. This continues to be posturing and very little else. Its nothing personal like I said it is just entirely too deceptive for me not to continue to comment on if you continue to make or imply those claims. You simply need to quit it. The only pointing out you are doing is to your self as the alleged expert and guru (while coincidentally selling your product on the same basis) not to even a single piece of independently verifiable evidence..

As someone pointed out, I am not obliged to do this. I am just saying as it happenes to me, and have repeatedly requested readers of this thread not to buy anything from me based on this thread.

Just because I sell my expert services does not mean I should not try to help others.

And Mike, you have repeated your thing enough number of times and emphasized your point well enough - and people who want to really read thins thread would surely find your viewpoint. I fully respect the fact that you are entitled to your viewpoint.

But, would you please refrain from repeating this same point yet again unless you have anything further to add to your substance? Please treat this as a Warrior's request to another Warrior. I am sure other Warriors too would want to read and agree/disagree with my beliefs - so how about leaving some leeway for others to express enough of their thoughts too while keeping the main content (good or bad whatever) readable and easy-to-follow still?

[jason_simpson]

Hey Fred, thanks for the lovely discussion here about SEO. I really enjoyed reading every single bit of it. This stuff is really knowledgeable.

[Mike Anthony]

Fred theres a simple way for me to stop posting responses pointing out that you have no inside secret to Google's algo. Stop making claims that you do.

Anyone reading this thread has a good idea who you are referring to with

"Some people here fail to appreciate"

So why refer to us and then complain when I respond? As long as you make references to me (even if among others) its perfectly with the rules of this board to respond.

I would never have posted in his thread again if you were not right back to implying that again and referring to those of us who don't accept what you have said. I've asked you to stop making the claim because its false. No one knows the precise mathematical equations Google uses and there you were again implying that other people just don't get your secret and taking a back hand swipe at what I fail to appreciate or grasp.

Certainly you are not required to post any evidence.Thats your choice and its my choice to point out that you have presented none. this isn't your blog. Its an open forum. IF you want to help people as you claim thats where the evidence IS required. Nothing does more harm to people than refusing to post any evidence but making extraordinary claims of being able to rank number one for almost anything. thats how people get scammed. Bogus claims minus any evidence.

[FredJones]

Mike, sorry but by saying things like: "no one understands" some basic SEO science I seriously think you have starte under-estimating the rest of the entire world.

What is your take on this material publcly available, by Page and Brin?

The Anatomy of a Search Engine

This also talks about the same things that I do, and just like I never try to imply absolute weightage on factors in my original posts beyond a certain extent, nor does thse.

Please stop making absurd claims. And please stop calling others liers as you are telling me to - I take it as an outright offense now. Just stop it.

And you had already talked about giving out the literature on "historical patent" which I am yet to see in the last 7 days - were you serious or were you kidding with all of us Warriors?

EDIT: Also, notice the reference of "Chakrabarti" in the references - this is Soumen Chakrabarti that I had referred to earlier and Page and Brin do happen to attribute some of core Google architecture to Soumen. In other words, just like we believe Page and Brin to be search engine authorities, they think Chakrabarti to be an authority.

Originally Posted by Mike Anthony

Fred theres a simple way for me to stop posting responses pointing out that you have no inside secret to Google's algo. Stop making claims that you do.

Anyone reading this thread has a good idea who you are referring to with

"Some people here fail to appreciate"

So why refer to us and then complain when I respond?

I would never have posted in his thread again if you were not right back to implying that again. I've asked you to stop making the claim because its false. No one knows the precise mathematical equations Google uses and there you were again implying that other people just don't get your secret and taking a back hand swipe at what I fail to appreciate or grasp.

Certainly you are not required to post any evidence.Thats your choice and its my choice to point out that you have presented none. this isn't your blog. Its an open forum. IF you want to help people as you claim thats where the evidence IS required. Nothing does more harm to people than refusing to post any evidence but making extraordinary claims of being able to rank number one for almost anything. thats how people get scammed. Bogus claims minus any evidence.

[Mike Anthony]

Originally Posted by FredJones

Mike, sorry but by saying things like: "no one understands" some basic SEO science I seriously think you have starte under-estimating the rest of the entire world.

You are addressing me directly by name this time so I think I can respond?

Its no one knows what is in the algo and it is a fact.

What is your take on this material publcly available, by Page and Brin?

The Anatomy of a Search Engine

This also talks about the same things that I do

It does not. It states nothing about a calculation regarding serps and it makes no claim that anyone can rank almost any time for almost any term. It has an example calculation on PR and thats about it. Using that as a pretext to make up other calculations is of no use. We know of pagerank because GOOGLE (critical point) has given us that. Its one sliver of the pie and knowing that alogo by itself will not rank you on Google or give you any superior knowledge of serps.

If you want your calculations to stand then you need to show where you got them from. IF they are from Google then yes I will take that as evidence. Fair enough? Can we have the link to where Google has your calculations?

Please stop making absurd claims.

I've made no absurd claim. I've asked you to stop making what I think anyone who knows SEO in this forum would consider an absurd claim - That you can rank number one for almost any term in the search engines by applying your algos.

I am not calling you a liar. I am calling what you claim as deceptive because its a blanket statement without any evidence and the nature of such statements with the lack of evidence are by their nature deceptive especially to newbies.

You can take this personal but it isn't. You might be a stand up guy. We are talking SEO and having a pointed disagreement on principle. No need to take it beyond that. Emotional states have nothing to do with it.

[FredJones]

Originally Posted by Mike Anthony

You are addressing me directly by name this time so I think I can respond?

Its no one knows what is in the algo and it is a fact.

It does not. It states nothing about a calculation regarding serps and it makes no claim that anyone can rank almost any time for almost any term. It has an example calculation on PR and thats about it. Using that as a pretext to make up other calculations is of no use. We know of pagerank because GOOGLE (critical point) has given us that. Its one sliver of the pie and knowing that alogo by itself will not rank you on Google or give you any superior knowledge of serps.

No, it starts with page rank but goes ahead to talk about anchor text (which is what I refer as the setter of context - section 2.2), search prximity and visual presentations such as font size (section 2.3), talks about document indexing (TF-IDF factors - section 4.2.3), lexicon (which translates to main keyword and LSI in my understanding, section 4.2.4) and ultimately the ranking system in section 4.5.1 - this is what Google ranking is mostly about even today with some changes here and there.

And it all comes from a level over and above Matt Cutts - it comes from the Gods of Google, namely Larry Page and Sergei Brin.

And they add the system architecture drawing which addresses some of the most important components of Google as a search engine.

If this is not knowledge on search related to Google then nothing is.

And I believe, this discussion has finally managed to make this thread richer

[Mike Anthony]

Originally Posted by FredJones

And you had already talked about giving out the literature on "historical patent" which I am yet to see in the last 7 days - were you serious or were you kidding with all of us Warriors?

What in Sam's Hill are you talking about? Google's "historical patent" is well known among Professional SEOs. I wouldn't think I needed to provide it for someone with the credentials and expertise you claim. Me Kidding? Google it.


United States Patent Application: 0050071741


Break down on it here

SEOmoz | Google's Patent - Information Retrieval Based on Historical Data

I don't fudge or make up facts on SEO. I've linked to it in other threads. Didn't think I needed to here again but there you have it.

[Mike Anthony]

Originally Posted by FredJones

No, it starts with page rank but goes ahead to talk about anchor text............

absolutely none of which has been in any dispute. What has been in dispute are your calculations that you use and the ARE NOT in that report. Stop fudging. The only usefulness of that is for you to hide behind it and pretend that your calculations are in there and this in any way bails you out of the preposterous claim that you can rank number one for any search term.