The Developer Tutorials Forum SEO FAQ!

[BigAlReturns]

I feel it would be a good idea to have a thread with common answers to SEO problems, as often the same questions can be asked again and again, and it's nice to have a central resource of information.
I've got started with a few questions and answers, but hopefully this can become an interactive thread, with both questions and answers submitted and added to the FAQ itself. Looking forward to your ideas!

THE SEO FAQ

What is PR, and how important is it?

Pagerank, often abbreviated to PR, is a term you'll hear a great deal in the SEO world. Briefly put, it is a numerical value calculated by Google, which is a measure of the incoming links to a particular page. The more incoming links you have, and the higher the PR of these incoming links, the greater the PR of that particular page will be. It's important to realise that PR is calculated on a page by page, not site by site basis, so even if a site's homepage has high PR, if the page linking to you is low PR, the link will only pass a small amount of PR.


The actual formula for calculating PR is no mystery, as it was part of a published paper by the Google founders, Sergey Brin and Larry Page (for whom PageRank is named!). The formula is:

• PR(A) = (1-d) + d (PR(T1)/C(T1) + ... + PR(Tn)/C(Tn))

where PR(A) is the PageRank of a page A
PR(T1) is the PageRank of a page T1
C(T1) is the number of outgoing links from the page T1
d is a damping factor in the range 0 < d < 1, usually set to 0.85


This looks a bit daunting, but basically all it's saying is that the PR of page A, PR(A), is initially set at 0.15. We then add to that figure for each incoming link, taking into account the number of other links that page has. The basic idea to take from this, is that a link that passes lots of PR is one from a high PR page with few links.


Now it's important to understand that the PR you know and love, the green bar on the top of your browser is not a great indication of the actual PR of a particular page. Actual PR is calculated by Google daily, and can theoretically range from 0.15 to infinity for any given page. The PR you see in your browser is this value expressed logarithmically, as a proportion of the highest PR page on the web. So a PR3 page could range, for example, from 1000-10,000 actual PR units. The next level up, PR4, might be 10,000 to 100,000 units. These are obviously arbitrary figures I made up, but they show how inaccurate the green bar in your browser is. And it gets worse...the toolbar PR is only updated by Google every 3 months or so, so the value you see was not only an approximation to start with, but it's now a 3 month old approximation.


A common misconception is that PR can only be passed by external links, but internal links are just as important. Managing the flow of PR through your site can be of great help in improving rankings.


You might have been able to guess by now, but my final sentiment is that PR is pretty much irrelevant in how well a page ranks. There are hundreds of factors that affect a page's ranking, and PR, in it's raw form is a very very small part of this.


What is duplicate content, what are its consequences and how do I prevent it?

Duplicate content is pretty much what it says on the tin, content which is so similar that it is deemed, by a search engine, to be an exact duplicate. The measures used by search engines to determine this are unknown, but it's best to take a common sense approach. If on reading the 2 pages a human would think they were copied from one to the other, they're likely to be seen as duplicates.

What are the consequences of duplicate content? Well, Google have stated that if they come across duplicate pages, they will make a judgement as to which is the “original,” and which is the “duplicate.” How exactly they do this we don't know, but it's likely to be based on when they first came across the page, and the number/quality of incoming links to each page. The original will be indexed and ranked as normal, with no penalties, the duplicate will be put in the supplemental index, and will stand no chance of ranking.

It's obviously important to ensure that our pages don't end up in the supplemental index, so how can we stop this. The first step is to make sure we don't rip content from other pages, and don't let other page use our content. If another page isn't indexed, you could theoretically use their content, and get away with being original, but to me this isn't morally acceptable, and is likely to breach copyright laws.

The next step is to ensure we don't have duplicate content issues within our own site. This can commonly occur when dynamically generating pages on a database driven site. While Google will probably index one of these pages, without penalty, the problem comes when we think of incoming links. If I have 5 duplicate pages, then people could link to any of these 5. However, only the links to the “original” will be counted by Google, so we're missing out on 80% of our incoming links, and this is with only 5 pages! The key is to follow good design principles, and use rewritten URLs and redirects where necessary.


Should I use absolute or relative URLs?


Absolute URLs (http://www.mysite.com/page1) have no implicit advantage over relative URLs (/page1) in terms of SEO, but they are still the preferred option, for two reasons.

1) At some point in the future the page on which the link appears may change location. If you use relative URLs the link will now be wrong, whereas with absolute URLs it will still work.
2) It's unfortunate, but for content writers there is always a worry of having your content scraped. Using absolute URLs means that not only are the links on the scraped content not pointing at the scraper's pages, they are in fact going to be incoming links to your page! This is hallmark of scraped pages, and SEs undoubtedly use this to differentiate between original and scraped content.

[BigAlReturns]

Will I be penalised for using “invisible” content?

The answer to this question depends entirely on what context you use it in. Using CSS to hide content has many legitimate uses, for example, a “Show More” link that expands a div to make more readable. In this case, the hiding is used to make a genuine usability improvement for the end user, and would be perfectly acceptable.

When hidden content is not acceptable is when it is used with the express purpose of increasing rankings. For example, having a list of keywords related to the page to increase relevance, which is hidden so as not to make the page look ugly – this is not acceptable to the search engines, and if caught, you can expect to be penalised.

As a simple rule, if the hidden content can be revealed by a user's action, such as a click or a rollover, then it's OK, if it can never be seen by the user, its not OK.


How important are meta keywords and descriptions to the search engines?

The short answer to this question is that they are of miniscule importance. The meta tags are never displayed to the end-user, and as such, a webmaster can manipulate them to say whatever he/she likes without affecting the user experience. This means that there is absolutely no reason why they can be assumed, by a search engine, to bear any relevance to the page itself.

Using this logic, meta keywords and descriptions are barely worth the trouble of writing, and this is indeed true for the keywords. However, there is an important reason why meta descriptions should be used. When Google lists your site in its search results, it attempts to choose a relevant slice of content to describe the page, and display. In certain cases, encountered reasonably often, there is no content on the page Google determines to be relevant to the query, and in these cases, the meta description is used instead. As such, it can act as your “advertising copy” within search results, and can and will have an influence on click-through rates to your site.

What is the optimum keyword density to use in my content?

This is one of my biggest bug-bears in SEO, to see people trying to conform to a so called “ideal” keyword density. Simply put, the keyword density is the number of occurences of a specific keyword, divided by the number of words on a page. Several numbers have been thrown around as ideals, ranging from 2-10% that I've seen. The thinking SEO, however, should recognise that there is little value in trying to match content to a given ratio. Your content is written for humans, and it is humans who will link to your pages. Write your content for humans, and not only will it read better, and give an improved user experience, but it will also be more likely to attract incoming links from other webmasters. These incoming links will give you a bigger ranking boost than any specific keyword density ever will.

[BigAlReturns]

What is/isn't a good link?

There are several criteria for what makes a good incoming link:

1. Link will not contain rel=”nofollow”
2. Page containing link ranks well for your targetted keywords
3. Page containing link has content related to yours.
4. Anchor text of link relates to your targetted keywords
5. Page containing link has high PR
6. Page containing link will send many visitors

I would say they are more or less in order of importance. Obviously not having nofollow is most important, otherwise the link might as well not exist in Google's eyes! Points 2,3 and 4 are very important. Google determines the relevance of a page to a given query in two ways – the content of that page, and what its incoming links say about its content. A link which scores well in all these areas will tell Google that your page is related to that keyword, improving your rankings.

PR ranks down below the related content points because a high PR, yet unrelated link will have little effect on your rankings. It could have an indirect effect, by increased PR => more pages indexed => better rankings, but unless it's a mega PR (8,9,10) link, this effect isn't likely to be huge.

The traffic issue is a small one for SEO purposes. More visitors means more potential people who could link to your site, but this is unlikely to have a big effect, unless your site is focussed towards many people who run websites. In the grand scheme of things, where traffic leads to revenue, a high traffic link may be one of your most important factors, but we're talking SEO exclusively here.

What isn't a good link is also important. A key point to remember is that no incoming link will ever reduce your rankings. If this were your case then your competitors would quickly set up spam sites and link to you from them. However, linkbuilding tactics that require you to put outgoing links on your site can and will have a negative effect, if used in an incorrect manner. The following link patterns are discouraged by Google, and will negatively impact your rankings:

1. Reciprocal linking
2. Link farms (although only if you have outgoings from your site)

[BigAlReturns]

Do TLDs (.com/.net/.org/etc.) matter?

A simple one this – no! (OK of course there are exceptions, but I'll go into those!) There are two contexts in which the question could be asked, firstly in terms of your own site, and secondly in terms of incoming links. Generally, people want to know if a link from a .gov or .edu site will give an advantage, because search engines “trust” .gov or .edu sites. While it seems like a nice idea, why should a SE give an arbitrary boost to a site, simply based on its domain name. If the site is of high quality, it will achieve good rankings by the means of incoming links, if it isn't, then it won't. The .gov/.edu is no stamp of quality! The confusion arises because .gov and .edu sites often are of high quality, and are well trusted by search engines. This, however, is purely because they contain great content, so people do link to them, so they do achieve the great rankings. The proof of the pudding is in the links, and it's because of these that the sites rank well, not because of any inherent advantage to being a .gov.

Now on to those pesky exceptions...the TLD can be important if you want to rank well in local searches. It's one of several factors used to determine if a site is relevant to a localised query, but that's a whole other issue for another question!

[BigAlReturns]

Good SEO Resources

• Submit to Google, Yahoo and MSN
• WordTracker - Great keyword research tool - costs money, but even the free trial is useful
• Google Webmaster Central - some great free tools to track performance
• SEOMoz Page Strength - a free tool which provides a metric as to your site's SEO performance. Also a good blog on this site.

Further Reading
SEO is constantly changing, and being good at it means staying up to date. I don't always agree with the authors of these sites on every issue, but learning is about taking in all you can, and making your own judgements.

• SEOMoz Blog
• SEOBook Blog (decent blog - but I wouldn't recommend the book!)
• Matt Cutts' (Google Engineer) Blog
• Blue Hat SEO - a look into the murkier, less moral methods of SEO
• ThreadWatch - interesting threads from many forums on the subject of SEO
• Google Webmaster Central Blog - not updated too often, but Google's main way of releasing official information

[BigAlReturns]

OK guys I've got some time this week to hopefully write a bit more in the FAQ - are there any questions you'd like to see answered, or suggestions for some topics? Either let me know your questions on here, or write the answers yourself and we'll stick it in the FAQ!