SEO Technique: Optimizing through copy blocks

Posted by Michael Martinez on October 11, 2007 in Search Engine Optimization

I don’t often share specific SEO tips and techniques because their usefulness doesn’t last very long, they may not work for everyone, and people sometimes take my ideas and turn them into spam techniques. While I cannot stop anyone from scripting specific ideas, the search engines do look for emerging patterns in Web page optimization. The people who need to take that warning most to heart are the people who do it by hand.

On-page optimization can be done largely by feel or “by the numbers”. People who do it by the numbers tend to produce formulaic, clunky copy. I write plenty of formulaic, clunky copy myself but I don’t particularly like it. You can shoot the goose that lays the golden eggs simply by writing copy that is so uncompelling that no one who lands on a page answers your call to action.


Point 1: Design Copy In Blocks


This is the hardest concept for people to grasp, although it seems as plain as day to me. Every Web page can be divided into blocks or sections. Some people call them “zones” (and I used to) but advertising networks now use their own zone layouts and I’m not referring to those fabricated zone maps.

A copy block is any section of a Web page where the copy makes one point and only one point. If you write well then you break up your copy so that each paragraph makes only one point, each caption makes only one point, etc. The length of text is irrelevant. The meaning of the text is everything. If you have a lengthy section of text that makes multiple points, you need to break up your text (at least until you are ready to put the page back together).

Copy blocks appear in mastheads, navigational tables, margin columns, main body columns, and in footers. There is no right or wrong place for copy blocks. You should be able to list all the copy blocks in your basic Web template.


Point 2: Assign keywords by copy blocks


There is an old SEO myth that suggests you should optimize a Web page for no more than 3-5 phrases, and the fewer the better. That is just absolute nonsense. You want to optimize your Web page for as many expressions as you possibly can without destroying the page’s aesthetic and functionality.

Pages are relevant to every word that occurs on them (some exceptions occur but for this article we’ll assume there are none). Pages are also relevant to every possible combination of words they contain. Proximity, placement, rareness, emphasis, and repetition all help make pages more relevant to specific expressions.

Proximity refers to how close the words in an expression happen to be to each other. It also implies that order is important. A page that contains “these words as quoted” is more relevant to that quoted expression than a page that contains “quoted as words these”, but both pages are relevant to either expression.

Placement refers to your use of words in specific page elements: titles, URLs, Hx tags, ALT= attributes, outbound link anchor text, etc. Think of placement as a different way to emphasize a word’s value to a document.

Rareness refers to how rare a word is deemed to be with respect to the universal lexicon, the Web as a whole. If you use “canine” as often as “dog”, your page will be more relevant to “canine” because “dog” occurs in more Web documents. For example, in Google’s hidden Supplemental Index, some very common words apparently are not being index, whereas most rare words do seem to be.

Rareness also determines how large the base document set really is. A word found only in 1,000 documents is more likely to help your page be scored highly relevant for a query than a word found in 1,000,000 documents. Some people use repetition to compensate for light rareness.

Emphasis is exactly what it sounds like. However you make a word stand out on the page (by bolding, italicizing, “quoting”, or otherwise giving the word special emphasis), the more prominently featured a word is compared to other words on the page, the more relevant that page is likely to be for that word.

Repetition is the spammer’s tool but it is also the Best Practice that is most often used in natural large copy pages (like blog posts with comments and forum posts). A page that is highly relevant to a common word may have that word repeated many times. I have counted legitimate repetitions of words in the hundreds of occurrences on one page.

Just because a word is used on a page 100 times does not mean it is spam. If you write a blog post and 100 people post lengthy comments, using your name at the top of each comment, your name will occur on the page at least 100 times. That’s natural. It can also be faked, so don’t assume that you have a template for spam in my example.


Point 3: Scoring for relevance


Each search engine decides for itself which of the above factors it will score for relevance. Each search engine decides for itself how each factor will be scored. You cannot affect that relevance scoring algorithm. That is sadly one of the most difficult lessons people to learn in this business.

You cannot second-guess a search engine but you can devise your own scoring system. You can use that scoring system to teach yourself to write pretty good copy that is highly optimized for many expressions.

Most people assume their relevance scores are driven by three primary factors: page titles, page URLs, and inbound link anchor text. Of those three factors, you can only rely on inbound link anchor text for heavy repetition. However, if you focus on the other on-page factors, you will find that repetition dove-tails nicely with copy blocks.

You should score each of your copy blocks for relevance to each block’s targeted expression. Each block can only target one expression optimally. It doesn’t matter how your score your copy block optimization as long as you are consistent. This scoring is for your own use. It’s not a predictor of how a search engine will evaluate the copy.

By looking at the mechanics of calculating relevance scores you’ll see how to write optimal copy for each block. You can, if you wish, dedicate more than one block of text to a specific expression, but you can also target many different expressions.


Point 4: Evaluating page relevance


Most queries are resolved on the basis of relevance scores. Although the search engines can and do weight results by “importance”, it has been shown time and time again that so-called “highly important” pages can be outranked in search results by less important pages.

Your objective in using your own page relevance scoring algorithm is to help you improve your writing skill. It’s like training with a wooden sword before you switch to a metal swod. The more you practice writing highly relevant copy, the more you create highly relevant copy. As you become comfortable with writing highly relevant copy, you can focus on other aspects of your writing, such as embedding appropriate calls to action, voice, tone, etc.

This methodology is not a standaloine technique that helps you achieve a high ranking quickly. This is a training method to help you improve your copy writing skill. You want to intuitively write highly relevant copy that generates many conversions (regardless of how you measure conversions).

I can say more about copy blocks but I think I’ll leave the rest for a later post.

8 Comments on SEO Technique: Optimizing through copy blocks

By wibbler on October 11, 2007 at 7:18 am

Hi there,

“A word found only in 1,000 documents is more likely to help your page be scored highly relevant for a query than a word found in 1,000,000 documents.”

Okay - lets say I had to write a page which I needed to rank for the word “astronomy”.
At 60,000,000 returns on google, this is a word which I would regard as very common in the web as a whole. In order to help my page become *more* relevant to this word (trust, PR, links, domain age excluded) *relative* to other pages which are currently ranking, must I also include symantically related rare(r) words on my page, such as -
celestial - 60,000,000
messier - 3,000,000
extrasolar - 2,000,000

The above is my interpretation of your statement quoted above. If my interpretation is not correct, then the only other thing I can think of which pertains to the quote in terms of helping my page, would be to write my page about a rarer topic - which would defeat the object though because I needed to rank for “astronomy”.

Am I thinking straight here?
Cheers
Wibbler

By Michael Martinez on October 11, 2007 at 8:23 am

Wibbler “Okay - lets say I had to write a page which I needed to rank for the word ‘astronomy’.
At 60,000,000 returns on google, this is a word which I would regard as very common in the web as a whole. In order to help my page become *more* relevant to this word (trust, PR, links, domain age excluded) *relative* to other pages which are currently ranking, must I also include symantically related rare(r) words on my page, such as”

Michael Not exactly. The fewer documents that use a word, the higher the probability that your use of the word will outscore other documents’ use of the word. Adding other words doesn’t make your document more relevant to the first word, but it does make your document more relevant to expressions using that word.

Hence, if you just use “astronomy” your document is competing with a lot of other documents. If you add the words “class” and “teacher” your document becomes relevant to “astronomy class” and “astronomy teacher”. On the other hand, if you use “messier” but don’t use the word “transformation”, your document may still be considered relevant for “messier transformation” if the search engine cannot find enough documents that use “messier transformation”.

Which is a tangent. My point is that relevance is in part relative to the number of documents that are relevant to a word. I don’t think it’s productive for people to pick out rare keywords in the hope of boosting relevance. It’s more productive, in my opinion, to write copy that is relevant to the term and then emphasize the term and repeat the term as much as is naturally acceptable. The rule of thumb is, “What does the human eye prefer?”.

But that’s a whole ‘nother post.

By wibbler on October 14, 2007 at 3:34 am

Ok - having read the reply (thanks very much for all your replies) - i’ll bear in mind what youve said there.

Its a tough one though still, because whilst ranking for lesser (but related) terms, is not always an option for me as brand is brand or product is product, and where there are 10,000 searches daily for “product” - there are only 10 for “product review”.

I certainly couldnt survive with only the 10 - so i am often pinned to targetting the words where there are already millions of pages containing it. (word or phrase that is).

Still reading.
Cheers
Wibbler

By wibbler on October 14, 2007 at 3:37 am

PS - I forgot to add that I have used WT for years to do my keyword research - and it is there where I find the info on subphrases to target where there are any - I never ignore the possibility of finding good extra phrases to include.

Cheers

By Michael Martinez on October 14, 2007 at 11:37 pm

Keep in mind that the shorter a query is, and the more competition there is for it, the more likely those high daily search numbers are inflated by rank-checking and so-called “competitive intelligence”.

That said, if you only get a slice of a very large pie, that can be pretty lucrative in some markets. You don’t want to use copy blocks to chase unrelated expressions. You want to use them to strengthen the quality of your copy with respect to the primary topic.

By dodito on October 19, 2007 at 8:22 am

Hmmm that leads us back to an old discussion: how much page optimization can you do for a word like “astronomy” to rank well, even when using an internal linking structure, nice copy blocks etc etc. Even when using every trick in the book and finding many “astronomy-other-word combinations” so you get the word “astronomy” 100 or more times on that same page, using url’s, page title, have that page in the top menu, even have it linked out of other pages of text .. then what is there left to do.. and.. don’t you need to resort to linkbuilding at some point to push the page ?

By dodito on October 19, 2007 at 8:25 am

I guess an alternative would be, to find completely different keywords that are related to your “astronomy” pages, and even find different related topics for which it is easier to score, but where you know people will be interested in astronomy as well (or your point of view/product/subcategory etc). Which then means… you can slowly build a presence which is able to “attack head on” the single key-word rankings ?

By Michael Martinez on October 19, 2007 at 1:07 pm

Some blog posts end up with 100-500 comments that repeat certain words over and over again. Those posts tend to rank very well in otherwise link-competitive queries. I cannot show you any examples for confidentiality reasons. Whether you associate your single keywords with other targeted keywords is not really germaine to the concept of copy blocks.

Each comment in this blog post is a copy block. We have used the word “astronomy” several times and each occurrence of the word “astronomy” in our comments makes the entire post more relevant to the word “astronomy”.

I wouldn’t be surprised to see people asking their visitors to spam their comments with specific keywords. The search engines might not be able to filter that very easily in the short run, but such short-sighted spamming would definitely be inadvisable.

For example, there are dozens — perhaps hundreds or even thousands — of blogs that are link spamming right now through a new tactic that requires them to replicate the words “This is based on the SEO theory that links to posts inside your blog are more important than links to your home page”. I’ve used that language twice on SEO Theory as a subtle sign to Googlers and other search engineers who read SEO Theory that there is something up.

But let me be a little more blatant: Googlers and other search engineers, you need to filter out the blogs that are using the language “This is based on the SEO theory that links to posts inside your blog are more important than links to your home page” in association with other obvious features that leave a very distinctive spam (link manipulative) footprint.

To all the bloggers who have used this link spam tactic: I suggest you delete those posts, because I’m going to keep complaining until something is done.

So I don’t advise anyone to ask their blog commenters to help spam posts to the top of search results with added comments, either. I’m just trying to illustrate how a Web page can naturally use a word hundreds of times without being spammy.

Comment

Log in or Register to post a comment.

More

Read more posts by Michael Martinez

About the Author

Michael Martinez is the Director of Search Strategies for Visible Technologies, Inc. A former moderator at SEO forums such as JimWorld an Spider-food, Michael has been active in search engine optimization since 1998 and Web site design and promotion since 1996. Michael was a regular contributor to Suite101 (1998-2003) and SEOmoz (2006).

Advanced SEO: The 4-step campaign process SEO Technique: Creating Copy Blocks