List of Best and Worst practices for designing a high traffic website
Thank you slashdot :-)
Here is a checklist of the factors that affect your rankings with Google, Bing, Yahoo! and the other search engines. The list contains positive, negative and neutral factors because all of them exist. Most of the factors in the checklist apply mainly to Google and partially to Bing, Yahoo! and all the other search engines of lesser importance. If you need more information on particular sections of the checklist, you may want to read our SEO tutorial, which gives more detailed explanations of Keywords, Links, Metatags, Visual Extras, etc.
Keywords in <title> tag
This is one of the most important places to have a keyword because what is written inside the <title> tag shows in search results as your page title. The title tag must be short (6 or 7 words at most) and the the keyword must be near the beginning.
Keywords in URL
Keywords in URLs help a lot - e.g. - http://domainname.com/seo-services.html, where "SEO services" is the keyword phrase you attempt to rank well for. But if you don't have the keywords in other parts of the document, don't rely on having them in the URL.
Keyword density in document text
Another very important factor you need to check. 3-7 % for major keywords is best, 1-2 for minor. Keyword density of over 10% is suspicious and looks more like keyword stuffing, than a naturally written text.
Keywords in anchor text
Also very important, especially for the anchor text of inbound links, because if you have the keyword in the anchor text in a link from another site, this is regarded as getting a vote from this site not only about your site in general, but about the keyword in particular.
Keywords in headings (<H1>, <H2>, etc. tags)
One more place where keywords count a lot. But beware that your page has actual text about the particular keyword.
Keywords in the beginning of a document
Also counts, though not as much as anchor text, title tag or headings. However, have in mind that the beginning of a document does not necessarily mean the first paragraph – for instance if you use tables, the first paragraph of text might be in the second half of the table.
Keywords in <alt> tags
Spiders don't read images but they do read their textual descriptions in the <alt> tag, so if you have images on your page, fill in the <alt> tag with some keywords about them.
Keywords in metatags
Less and less important, especially for Google. Yahoo! and Bing still rely on them, so if you are optimizing for Yahoo! or Bing, fill these tags properly. In any case, filling these tags properly will not hurt, so do it.
Keyword proximity measures how close in the text the keywords are. It is best if they are immediately one after the other (e.g. "dog food"), with no other words between them. For instance, if you have "dog" in the first paragraph and "food" in the third paragraph, this also counts but not as much as having the phrase "dog food" without any other words in between. Keyword proximity is applicable for keyword phrases that consist of 2 or more words.
In addition to keywords, you can optimize for keyword phrases that consist of several words – e.g. "SEO services". It is best when the keyword phrases you optimize for are popular ones, so you can get a lot of exact matches of the search string but sometimes it makes sense to optimize for 2 or 3 separate keywords ("SEO" and "services") than for one phrase that might occasionally get an exact match.
Optimizing for secondary keywords can be a golden mine because when everybody else is optimizing for the most popular keywords, there will be less competition (and probably more hits) for pages that are optimized for the minor words. For instance, "real estate new jersey" might have thousand times less hits than "real estate" only but if you are operating in New Jersey, you will get less but considerably better targeted traffic.
For English this is not so much of a factor because words that stem from the same root (e.g. dog, dogs, doggy, etc.) are considered related and if you have "dog" on your page, you will get hits for "dogs" and "doggy" as well, but for other languages keywords stemming could be an issue because different words that stem from the same root are considered as not related and you might need to optimize for all of them.
Optimizing for synonyms of the target keywords, in addition to the main keywords. This is good for sites in English, for which search engines are smart enough to use synonyms as well, when ranking sites but for many other languages synonyms are not taken into account, when calculating rankings and relevancy.
Spelling errors are very frequent and if you know that your target keywords have popular misspellings or alternative spellings (i.e. Christmas and Xmas), you might be tempted to optimize for them. Yes, this might get you some more traffic but having spelling mistakes on your site does not make a good impression, so you'd better don't do it, or do it only in the metatags.
When you are optimizing for an excessive amount of keywords, especially unrelated ones, this will affect the performance of all your keywords and even the major ones will be lost (diluted) in the text.
Any artificially inflated keyword density (10% and over) is keyword stuffing and you risk getting banned from search engines.
Links - internal, inbound, outbound
Anchor text of inbound links
As discussed in the Keywords section, this is one of the most important factors for good rankings. It is best if you have a keyword in the anchor text but even if you don't, it is still OK.
Origin of inbound links
Besides the anchor text, it is important if the site that links to you is a reputable one or not. Generally sites with greater Google PR are considered reputable.
Links from similar sites
Having links from similar sites is very, very useful. It indicates that the competition is voting for you and you are popular within your topical community.
Links from .edu and .gov sites
These links are precious because .edu and .gov sites are more reputable than .com. .biz, .info, etc. domains. Additionally, such links are hard to obtain.
Generally the more, the better. But the reputation of the sites that link to you is more important than their number. Also important is their anchor text, is there a keyword in it, how old are they, etc.
Anchor text of internal links
This also matters, though not as much as the anchor text of inbound links.
The text that is immediately before and after the anchor text also matters because it further indicates the relevance of the link – i.e. if the link is artificial or it naturally flows in the text.
Age of inbound links
The older, the better. Getting many new links in a short time suggests buying them.
Links from directories
Great, though it strongly depends on which directories. Being listed in DMOZ, Yahoo Directory and similar directories is a great boost for your ranking but having tons of links from PR0 directories is useless and it can even be regarded as link spamming, if you have hundreds or thousands of such links.
Number of outgoing links on the page that links to you
The fewer, the better for you because this way your link looks more important.
Named anchors (the target place of internal links) are useful for internal navigation but are also useful for SEO because you stress additionally that a particular page, paragraph or text is important. In the code, named anchors look like this: <A href= "#dogs">Read about dogs</A> and "#dogs" is the named anchor.
IP address of inbound link
Google denies that they discriminate against links that come from the same IP address or C class of addresses, so for Google the IP address can be considered neutral to the weight of inbound links. However, Bing and Yahoo! may discard links from the same IPs or IP classes, so it is always better to get links from different IPs.
Inbound links from link farms and other suspicious sites
This does not affect you in any way, provided that the links are not reciprocal. The idea is that it is beyond your control to define what a link farm links to, so you don't get penalized when such sites link to you because this is not your fault but in any case you'd better stay away from link farms and similar suspicious sites.
Many outgoing links
Google does not like pages that consists mainly of links, so you'd better keep them under 100 per page. Having many outgoing links does not get you any benefits in terms of ranking and could even make your situation worse.
Excessive linking, link spamming
It is bad for your rankings, when you have many links to/from the same sites (even if it is not a cross- linking scheme or links to bad neighbors) because it suggests link buying or at least spamming. In the best case only some of the links are taken into account for SEO rankings.
Outbound links to link farms and other suspicious sites
Unlike inbound links from link farms and other suspicious sites, outbound links to bad neighbors can drown you. You need periodically to check the status of the sites you link to because sometimes good sites become bad neighbors and vice versa.
Cross-linking occurs when site A links to site B, site B links to site C and site C links back to site A. This is the simplest example but more complex schemes are possible. Cross-linking looks like disguised reciprocal link trading and is penalized.
Single pixel links
when you have a link that is a pixel or so wide it is invisible for humans, so nobody will click on it and it is obvious that this link is an attempt to manipulate search engines.
Metatags are becoming less and less important but if there are metatags that still matter, these are the <description> and <keywords> ones. Use the <Description> metatag to write the description of your site. Besides the fact that metatags still rock on Bing and Yahoo!, the <Description> metatag has one more advantage – it sometimes pops in the description of your site in search results.
The <Keywords> metatag also matters, though as all metatags it gets almost no attention from Google and some attention from Bing and Yahoo! Keep the metatag reasonably long – 10 to 20 keywords at most. Don't stuff the <Keywords> tag with keywords that you don't have on the page, this is bad for your rankings.
If your site is language-specific, don't leave this tag empty. Search engines have more sophisticated ways of determining the language of a page than relying on the <language>metatag but they still consider it.
The <Refresh> metatag is one way to redirect visitors from your site to another. Only do it if you have recently migrated your site to a new domain and you need to temporarily redirect visitors. When used for a long time, the <refresh> metatag is regarded as unethical practice and this can hurt your ratings. In any case, redirecting through 301 is much better.
Having more content (relevant content, which is different from the content on other sites both in wording and topics) is a real boost for your site's rankings.
Frequency of content change
Frequent changes are favored. It is great when you constantly add new content but it is not so great when you only make small updates to existing content.
Keywords font size
When a keyword in the document text is in a larger font size in comparison to other on-page text, this makes it more noticeable, so therefore it is more important than the rest of the text. The same applies to headings (<h1>, <h2>, etc.), which generally are in larger font size than the rest of the text.
Bold and italic are another way to emphasize important words and phrases. However, use bold, italic and larger font sizes within reason because otherwise you might achieve just the opposite effect.
Age of document
Recent documents (or at least regularly updated ones) are favored.
Generally long pages are not favored, or at least you can achieve better rankings if you have 3 short rather than 1 long page on a given topic, so split long pages into multiple smaller ones.
From a marketing point of view content separation (based on IP, browser type, etc.) might be great but for SEO it is bad because when you have one URL and differing content, search engines get confused what the actual content of the page is.
Poor coding and design
Search engines say that they do not want poorly designed and coded sites, though there are hardly sites that are banned because of messy code or ugly images but when the design and/or coding of a site is poor, the site might not be indexable at all, so in this sense poor code and design can harm you a lot.
Using other people's copyrighted content without their permission or using content that promotes legal violations can get you kicked out of search engines.
This is a black hat SEO practice and when spiders discover that you have text specially for them but not for humans, don't be surprised by the penalty.
Cloaking is another illegal technique, which partially involves content separation because spiders see one page (highly-optimized, of course), and everybody else is presented with another version of the same page.
Creating pages that aim to trick spiders that your site is a highly-relevant one when it is not, is another way to get the kick from search engines.
When you have the same content on several pages on the site, this will not make your site look larger because the duplicate content penalty kicks in. To a lesser degree duplicate content applies to pages that reside on other sites but obviously these cases are not always banned – i.e. article directories or mirror sites do exist and prosper.
Visual Extras and SEO
Images in text
Having a text-only site is so boring but having many images and no text is a SEO sin. Always provide in the <alt> tag a meaningful description of an image but don't stuff it with keywords or irrelevant information.
Podcasts and videos
Podcasts and videos are becoming more and more popular but as with all non-textual goodies, search engines can't read them, so if you don't have the tapescript of the podcast or the video, it is as if the podcast or movie is not there because it will not be indexed by search engines.
Images instead of text links
Using images instead of text links is bad, especially when you don't fill in the <alt> tag. But even if you fill in the <alt> tag, it is not the same as having a bold, underlined, 16-pt. link, so use images for navigation only if this is really vital for the graphic layout of your site.
Frames are very, very bad for SEO. Avoid using them unless really necessary.
Spiders don't index the content of Flash movies, so if you use Flash on your site, don't forget to give it an alternative textual description.
A Flash home page
Fortunately this epidemic disease seems to have come to an end. Having a Flash home page (and sometimes whole sections of your site) and no HTML version, is a SEO suicide.
A very important factor, especially for Yahoo! and Bing.
Another fundamental issue, which that is often neglected. If the site (or separate pages) is unaccessible because of broken links, 404 errors, password-protected areas and other similar reasons, then the site simply can't be indexed.
It is great to have a complete and up-to-date sitemap, spiders love it, no matter if it is a plain old HTML sitemap or the special Google sitemap format.
Spiders love large sites, so generally it is the bigger, the better. However, big sites become user-unfriendly and difficult to navigate, so sometimes it makes sense to separate a big site into a couple of smaller ones. On the other hand, there are hardly sites that are penalized because they are 10,000+ pages, so don't split your size in pieces only because it is getting larger and larger.
Similarly to wine, older sites are respected more. The idea is that an old, established site is more trustworthy (they have been around and are here to stay) than a new site that has just poped up and might soon disappear.
It is not only keywords in URLs and on page that matter. The site theme is even more important for good ranking because when the site fits into one theme, this boosts the rankings of all its pages that are related to this theme.
File Location on Site
File location is important and files that are located in the root directory or near it tend to rank better than files that are buried 5 or more levels below.
Not all TLDs are equal. There are TLDs that are better than others. For instance, the most popular TLD – .com – is much better than .ws, .biz, or .info domains but (all equal) nothing beats an old .edu or .org domain.
Hyphens in URLs
Hyphens between the words in an URL increase readability and help with SEO rankings. This applies both to hyphens in domain names and in the rest of the URL.
Generally doesn't matter but if it is a very long URL-s, this starts to look spammy, so avoid having more than 10 words in the URL (3 or 4 for the domain name itself and 6 or 7 for the rest of address is acceptable).
Could matter only for shared hosting or when a site is hosted with a free hosting provider, when the IP or the whole C-class of IP addresses is blacklisted due to spamming or other illegal practices.
Adsense will boost your ranking
Adsense is not related in any way to SEO ranking. Google will definitely not give you a ranking bonus because of hosting Adsense ads. Adsense might boost your income but this has nothing to do with your search rankings.
Adwords will boost your ranking
Similarly to Adsense, Adwords has nothing to do with your search rankings. Adwords will bring more traffic to your site but this will not affect your rankings in whatsoever way.
Hosting downtime is directly related to accessibility because if a site is frequently down, it can't be indexed. But in practice this is a factor only if your hosting provider is really unreliable and has less than 97-98% uptime.
Spiders prefer static URLs, though you will see many dynamic pages on top positions. Long dynamic URLs (over 100 characters) are really bad and in any case you'd better use a tool torewrite dynamic URLs in something more human- and SEO-friendly.
This is even worse than dynamic URLs. Don't use session IDs for information that you'd like to be indexed by spiders.
Bans in robots.txt
If indexing of a considerable portion of the site is banned, this is likely to affect the nonbanned part as well because spiders will come less frequently to a "noindex" site.
Redirects (301 and 302)
When not applied properly, redirects can hurt a lot – the target page might not open, or worse – a redirect can be regarded as a black hat technique, when the visitor is immediately taken to a different page.