Chapter 6 Developing an SEO-Friendly Website

information architect

1) the individual who organizes the patterns inherent in data, making the complex clear. 2) a person who creates the structure or map of information which allows others to find their personal paths to knowledge. 3) the emerging 21st century professional occupation addressing the needs of the age focused upon clarity, human understanding, and the science of the organization of information.

To create your XML sitemap, you can use:

An XML sitemap generator: this is a simple script that you can configure to automatically create sitemaps, and sometimes submit them as well. Sitemap generators can create these sitemaps from a URL list, access logs, or a directory path hosting static files corresponding to URLs. Here are some examples of XML sitemap generators: -SourceForge.net's Google-sitemap_gen (http://bit.ly/google-sitemap_gen) -XML-Sitemaps.com Sitemap Generator (http://www.xml-sitemaps.com/) -Sitemaps Pal (http://www.sitemapspal.com/) -GSite Crawler (http://gsitecrawler.com/) Simple text: you can provide Google with a simple text file that contains one URL per line. However, Google recommends that once you have a text sitemap file for your site, you use the sitemap generator to create a sitemap from this text file using the Sitemaps protocol. Syndication feed: Google accepts Really Simply Syndication (RSS) 2.0 and Atom 1.0 feeds. Note that the feed may provide information on recent URLs only.

12 tips for selecting a great domain name:

Brainstorm five top keywords: when you begin your domain name search, it helps to have five terms or phrases in mind that best describe the domain you're seeking. Once you have this list, you can start to pair them or add prefixes and suffixes to create good domain ideas. For example, if you're launching a mortgage-related domain, you might start with words such as mortgage, finance, home equity, interest rate, and house payment, and then play around until you can find a good match. Make the domain unique: having your website confused with a popular site that someone else already owns is a recipe for disaster. Thus, never choose a domain that is simply the plural, hyphenated, or misspelled version of an already established domain. For example, for years Flickr did not own http://flicker.com, and the company probably lost traffic because of that. It recognized the problem and bought the domain, and as a result http://flicker.com now redirects to http://flickr.com Choose only dot-com-available domains: if you're not concerned with type-in traffic, branding, or name recognition, you don't need to worry about this one. However, if you're at all serious about building a successful website over the long term, you should be worried about all of these leements, and although directing traffic to a .net or .org (or any other new gTLDs) is fine, owning and 301-ing the .com or the ccTLD for the country your website serves (e.g., .co.uk for the United Kingdom), is critical. With the exception of the very tech-savvy, most people who use the Web still make the automatic assumption that .com is all that's out there, or that it's more trustworthy. Don't make the mistake of locking out or losing traffic from these folks. Make it easy to type: if a domain name requires considerable attention to type correctly due to spelling, length, or the use of unmemorable words or sounds, you've lost a good portion of your branding and marketing value. Usability folks even tout the value of having the words include easy-to-type letters (which we interpret as avoiding q, z, x, c, and p). Make it easy to remember: remember that word-of-mouth marketing relies on the ease with which the domain can be called to mind. You don't want to be the company with the terrific website that no one can ever remember to tell their friends about because they can't remember the domain name. Keep the name as short as possible: short names are easy to type and easy to remember. Short names also allow more of the URL to display in the SERPs and are a better fit on business cards and other offline media. Create and fulfill expectations: when someone hears about your domain name for the first time, he should be able to instantly and accurately guess the type of content he might find there. That's why we love domain names such as NYTimes.com, CareerBuilder.com, AutoTrader.com, and WebMD.com. Domains such as Monster.com, Amazon.com, and Zillow.com required far more branding because of their nonintuitive names. Avoid trademark infringement: this is a mistake that isn't made too often, but it can kill a great domain and a great company when it does. To be sure you're not infringing on anyone's registered trademark with your site's name, visit the U.S. Patent and Trademark office site (http://www.uspto.gov/trademarks/index.jsp) and search before you buy. Knowingly purchasing a domain with bad-faith intent that includes a trademarked term is a form of cybersquatting referred to as domain squatting. Set yourself apart with a brand: using a unique moniker is a great way to build additional value with your domain name. A "brand" is more than just a combination of words, which is why names such as Mortgageforyourhome.com and Shoesandboots.com aren't as compelling as branded names such as Yelp and Gilt. Reject hyphens and numbers: both hyphens and numbers make it hard to convey your domain name verbally and fall down on being easy to remember or type. Avoid spelled-out or Roman numerals in domains, as both can be confusing and mistaken for the other. Don't follow the latest trends: website names that rely on odd misspellings, multiple hyphens (such as the SEO-optimized domains of the early 2000s), or uninspiring short adjectives (such as top x, best x, and hot x) aren't always the best choice. This isn't a hard-and-fast rule, but in the world of naming conventions in general, if everyone else is doing it, that doesn't mean it is a surefire strategy. Just look at all the people who named their business AAA x over the past 50 years to be first in the phone book; how many Fortune 1000s are named AAA Company? Use a domain selection tool: websites such as Nameboy (http://www.nameboy.com) make it exceptionally easy to determine the availability of a domain name. Just remember that you don't have to buy through these services. You can find an available name that you like, and then go to your registrar of choice. You can also try BuyDomains (http://www.buydomains.com/) as an option to attempt to purchase domains that have already been registered.

Guidelines for selecting optimal URLs for the pages of your site(s):

Describe your content: an obvious URL is a great URL. If a user can look at the address bar (or a pasted link) and make an accurate guess about the content of the page before ever reaching it, you've done your job. These URLs get pasted, shared, emailed, written down, and yes, even recognized by the engines. Keep it short: Brevity is a virtue. The shorter the URL, the easier it is to copy and paste, read over the phone, write on a business card, or use in a hundred other unorthodox fashions, all of which spell better usability and increased branding. Remember, however, that you can always create a shortened URL for marketing purposes that redirects to the destination URL of your content - just know that this short URL will have no SEO value. Static is the way: Search engines treat static URLs differently than dynamic ones. Users also are not fond of URLs in which the big players are ?, θ, and =. They are just harder to read and understand. Descriptive text is better than numbers: if you're thinking of using 114/cat223/, you should go with /brand/adidas/ instead. Even if the descriptive text isn't a keyword or is not particularly informative to an uninitiated user, it is far better to use words when possible. If nothing else, your team members will thank you for making it that much easier to identify problems in development and testing. Keywords never hurt: if you know you're going to be targeting a lot of competitive keyword phrases on your website for search traffic, you'll want every advantage you can get. Keywords are certainly one element of that strategy, so take the list from marketing, map it to the proper pages, and get to work. For dynamically created pages through a CMS, create the option of including keywords in the URL. Subdomains aren't always the answer: first off, never use multiple subdomains (e.g., product.brand.site.com); they are unnecessarily complex and lengthy. Second, consider that subdomains have the potential to be treated separately from the primary domain when it comes to passing link and trust value. In most cases where just a few subdomains are used and there's good interlinking, it won't hurt, but be aware of the downsides. Fewer folders: a URL should contain no unnecessary folders (or words or characters, for that matter). They do not add to the user experience of the site and can in fact confuse users. Hyphens separate best: when creating URLs with multiple words in the format of a phrase, hyphens are best to separate the terms (e.g., /brands/dolce-and-gabbana/), but you can also use plus signs (+). Stick with conventions: if your site uses a single format throughout, don't consider making one section unique. Stick to your URL guidelines once they are established so that your users (and future site developers) will have a clear idea of how content is organized into folders and pages. This can apply globally as well as for sites that share platforms, brands, and so on. Don't be case-sensitive: URLs can accept both uppercase and lowercase characters, so don't ever, ever, allow any uppercase letters in your structure. Unix/Linux-based web servers are case-sensitive, so http://www.domain.com/Products/widgets/ is technically a different URL from http://www.domain.com/products/widgets/. Note that this is not true in Microsoft IIS servers, but there are a lot of Apache web servers out there. In addition, this is confusing to users, and potentially to search engine spiders as well. Google sees any URLs with even a single unique character as unique URLs. So if your site shows the same content on www.domain.com/Products/widgets/ and www.domain.com/products/widgets/, it could be seen as duplicate content. If you have such URLs now, implement a 301-redirect pointing them to all-lowercase versions, to help avoid confusion. If you have a lot of type-in traffic, you might even consider a 301 rule that sends any incorrect capitalization permutation to its rightful home. Don't append extraneous data: there is no point in having a URL exist in which removing characters generates the same content. You can be virtually assured that people on the Web will figure it out; link to you in different fashions; confuse themselves, their readers, and the search engines (with duplicate content issues); and then complain about it.

keyword cannibalization

a site's targeting of popular keyword search phrases on multiple pages, forcing the engines to pick which one is most relevant. A site employing cannibalization competes with itself for tanking and dilutes the ranking power of internal anchor text, external links, and keyword relevancy.

card sorting

a user-testing technique whereby users are asked to group items together so that you can organize your site as intuitively as possible. Card sorting can help identify not only the most logical paths through your site, but also ambiguous or cryptic terminology that should be reworded. With card sorting, you write all the major concepts onto a set of cards that are large enough for participants to read, manipulate, and organize. Your test group assembles the cards in the order they believe provides the most logical flow, as well as into groups that seem to fit together.

TF-IDF: two parts

term frequency: relates to the frequency of usage of a keyword or key phrase on a page, in comparison to usage levels in competing documents. Similar to keyword density, except weighting is done logarithmically to reduce the impact of keyword repetition. The result is that a page which uses a phrase 10 times might be seen only as twice as good a match as a page that uses that phrase once. Inverse document frequency (IDF): more about identifying uniqueness of a term. ex: the word omnipotent is used much less on the web than the word powerful. Therefore, a page using the word omnipotent may be seen as a bit more unique. If a user enters the word omnipotent as part of a search query, it will be far more likely to surface a page using that word in the results. IDF can be a very way to identify new ranking opportunities for your web page. TF-IDF helps search engines understand what terms a page emphasizes most, and what terms most uniquely define a page at the same time.

Pagination

the practice of creating a list of elements on pages separated solely by numbers

Guidelines for developing search engine-friendly navigation:

Implement a text-link-based navigational structure: if you choose to create navigation in Flash, JavaScript, or other technologies that the search engine may not be able to parse, make sure to offer alternative text links in HTML for spiders to ensure that automated robots (and visitors who may not have the required browser plug-ins) can reach your pages. Beware of "spider traps": even intelligently coded search engine spiders can get lost in infinite loops of links that pass between pages on a site. Intelligent architecture that avoids recursively looping 301 or 302 HTTP server codes (or other redirection protocols) should negate this issue, but sometimes online calendar links, infinite pagination that loops, or content being accessed or sorted in a multitude of ways (faceted navigation) can create tens of thousands of pages for search engine spiders when you intend to have only a few dozen true pages of content. Watch out for session IDs and cookies: if you limit a user's ability to view pages or redirect based on a cookie setting or session ID, search engines may be unable to crawl your content. The bots do not have cookies enabled, nor can they deal with session IDs properly (each visit by the crawler gets a URL with a different session ID, and the search engine sees these URLs with session IDs as different URLs). Although restricting form submissions is fine (as search spiders can't submit forms anyway), limiting content access via cookies and session IDs is a bad idea. Be mindful of server, hosting, and IP issues: server issues rarely cause search engine ranking problems - but when they do, disastrous consequences can follow. The engines are acutely aware of common server problems, such as downtime or overloading, and will give you the benefit of the doubt (though this will mean your content cannot be spidered during periods of server dysfunction). On the flip side, sites hosted on content delivery networks (CDNs) may get crawled more heavily, and CDNs offer significant performance enhancements to a website.

As search engines scour the Web, they identify four kinds of web structures on which to place metrics:

Individual pages/URLs: these are the most basic elements of the Web - filenames, much like those that have been found on computers for decades, which indicate unique documents. Search engines assign query-independent scores - most famously, Google's PageRank - to URLs and jusge them in their ranking algorithms. A typical URL might look something like http://www.yourdomain.com/page Subfolders: the folder structures that websites use can also inherit or be assigned metrics by search engines (though there's very little information to suggest that they are used one way or another). Luckily, they are an easy structure to understand. In the URL http://www.yourdomain.com/blog/post17, /blog/ is the subfolder and post 17 is the name of the file in that subfolder. Engines may identify common features of documents in a given subfolder and assign metrics to these (such as how frequently the content changes, how important these documents are in general, or how unique the content is that exists in these subfolders). Subdomains/fully qualified domains (FQDs)/third-level domains: In the URL http://blog.yourdomain.com/page, three kinds of domain levels are present. The top-level domain (also called TLD or domain extension) is .com, the second-level domain is yourdomain, and the third-level domain is blog. The third-level domain is sometimes referred to as a subdomain. Common web nomenclature does not typically apply the word subdomain when referring to www, although technically, this too is a subdomain. A fully qualified domain is the combination of the elements required to identify the location of the server where the content can be found (in this example, blog.yourdomain.com/). These structures can receive individual assignments of importance, trustworthiness, and value from the engines, independent of their second-level domains, particularly on hosted publishing platforms such as WordPress, Blogspot, and so on. Complete root domains/host domain/pay-level domains (PLDs)/second-level domains: the domain name you need to register and pay for, and the one you point DNS settings toward, is the second-level domain (though its commonly improperly called the "top-level" domain). In the URL http://www.yourdomain.com/page, yourdomain.com is the second-level domain. Other naming conventions may refer to this as the "root" or "pay-level" domain.

Reasons why pages may not be reachable by web crawlers:

Links in submission-required forms: search spiders will rarely, if ever, attempt to "submit" forms, and thus, any content or links that are accessible only via a form are invisible to the engines. This even applies to simple forms such as user logins, search boxes, or some types of pull-down lists. Links in hard-to-parse JavaScript: if you use JavaScript for links, you may find that search engines either do not crawl or give very little weight to the embedded links. In June 2014, Google announced enhanced crawling of JavaScript and CSS. Google can now render some JavaScript and follow some JaveScript links. Due to this change, Google recommends against blocking it from crawling your JavaScript and CSS files. For a preview of how your site might render according to Google, go to Search Console -> Crawl -> Fetch as Google, input the URL you would like to preview, and select "Fetch and Render." Links in Java or other plug-ins: Traditionally, links embedded inside Java and plug-ins have been invisible to the engines. Links in Flash: in theory, search engines can detect links within Flash, but don't rely too heavily on this capability. Links in PowerPoint and PDF files: search engines sometimes report links seen in PowerPoint files or PDFs. These links are believed to be counted the same as links embedded in HTML documents. Links pointing to pages blocked by the meta robots tag, rel="nofollow", or robots.txt: the robots.txt file provides a very simple means for preventing web spiders from crawling pages on your site. Using the nofollow attribute on a link, or placing the meta robots nofollow tag with the content="nofollow" attribute on the page containing the link, instructs the search engine to not pass link authority via the link. The effectiveness of the nofollow attribute on links has greatly diminished to the point of irrelevance as a result of overmanipulation by aggressive SEO practitioners. Links on pages with many hundreds or thousands of links: historically, Google had suggested a maximum of 100 links per page before it may stop spidering additional links from that page, but this recommendation has softened over time. Think of it more as a strategic guideline for passing PageRank. If a page has 200 links on it, then none of the links get very much PageRank. Managing how you pass PageRank by limiting the number of links is usually a good idea. Tools such as Screaming Frog (http://www.screamingfrog.co.uk/seo-spider/) can run reports on the number of outgoing links you have per page. Links in frames or iframes: technically, links in both frames and iframes can be crawled, but both present structural issues for the engines in terms of organization and following. Unless you're an advanced user with a good technical understanding of how search engines index and follow links in frames, it is best to stay away from them as a place to offer links for crawling purposes.

Best practices for <title> tag construction:

Place your keywords at the beginning of the <title> tag: this positioning provides the most search engine benefit; thus, if you want to employ your brand name in the <title> tag as well, place it at the end. There is a trade-off here, however, between SEO benefit and branding benefit that you should think about: major brands may want to place their brand at the start of the <title> tag, as it may increase click-through rates. To decide which way to go, you need to consider which need is greater for your business. Limit length to 50 characters (including spaces): content in <title> tags after 50 characters is probably given less weight by the search engines. In addition, the display of your <title> tag in the SERPs may get cut off as early as 49 characters. There is not hard-and-fast rule for how many characters Google will display. Google now truncates the display after a certain number of pixels, so the exact characters you use may vary in width. At the time of this writing, this width varies from 482px to 552px depending on your operating system and platform. Also be aware that Google may not use your <title> tag in the SERPs. Google frequently chooses to modify your <title> tag based on several different factors that are beyond your control. If this is happening to you, it may be an indication that Google thinks that your <title> tag does not accurately reflect the content of the page, and you should probably consider updating either your <title> tags or your content. Incorporate keyword phrases: this one may seem obvious, but it is critical to prominently include in your <title> tag the keywords your research shows as being the most valuable for capturing searches. Target longer phrases if they are relevant: when choosing what keywords to include in a <title> tag, use as many as are completely relevant to the page at hand while remaining accurate and descriptive. Thus, it can be much more valuable to have a <title> tag such as "SkiDudes | Downhill Skiing Equipment & Accessories" rather than simply "SKiDudes | Skiing Equipment." Including those additional terms that are both relevant to the page and receive significant search traffic can bolster your page's value. However, if you have separate landing pages for "skiing accessories" versus "skiing equipment," don't include one term in the other's title. You'll be cannibalizing your rankings by forcing the engines to choose which page on your site is more relevant for that phrase, and they might get it wrong. Use a divider: when you're splitting up the brand from the descriptive text, options include | (a.k.a. the pipe), >, -, and :, all of which work well. You can also combine these where appropriate ex: Major Brand Name: Product Category - Product. These characters do not bring an SEO benefit, but they can enhance the readability of your title. Focus on click-through and conversion rates: the <title> tag is exceptionally similar to the title you might write for paid search ads, only it is harder to measure and improve because the stats aren't provided for you as easily. However, if you target a market that is relatively stable in search volume week to week, you can do some testing with your <title> tags and improve the click-through rate. Watch your analytics and, if it makes sense, buy search ads on the page to test click-through and conversion rates of different ad text as well, even if it is for just a week or two. You can then look at those results and incorporate them into your titles, which can make a huge difference in the long run. A word of warning, through: don't focus entirely on click-through rates. Remember to continue measuring conversion rates. Target searcher intent: when writing titles for your web pages, keep in mind the search terms your audience employed to reach your site. If the intent is browsing or research-based, a more descriptive <title> tag is appropriate. If you're reasonably sure the intent is a purchase, download, or other action, make it clear in your title that this function can be performed at your site. ex <title> tag: PS4: PlayStation 4 Games & Consoles - Best Buy. The <title> tag here makes it clear that you can buy PS4 games and consoles at Best Buy. Communicate with human readers: this needs to remain a primary objective. Even as you follow the other rules here to create a <title> tag that is useful to the search engines, remember that humans will likely see your <title> tag presented in the search results for your page. Don't scare them away with a <title> tag that looks like it's written for a machine. Be consistent: Once you've determined a good formula for your pages in a given section or area of your site, stick to that regimen. You'll find that as you become a trusted and successful "brand" in the SERPs, users will seek out your pages on a subject area and have expectations that you'll want to fulfill.

Reasons to not use a microsite:

Search algorithms favor large, authoritative domains: take a piece of great content about a topic and toss it onto a small, mom-and-pop website; point some external links to it, optimize the page and the site for the target terms, and get it indexed. Now, take that exact same content and place it on Wikipedia, or CNN.com, and you're virtually guaranteed that the content on the large, authoritative domain will outrank the content on the small niche site. The engines' current algorithms favor sites that have built trust, authority, consistency, and history. Multiple sites split the benefits of links: a single good link pointing to a page on a domain positively influences the entire domain and every page on it. Because of this phenomenon, it is much more valuable to have any link you can possibly get pointing to the same domain to help boost the rank and value of the pages on it. Having content or keyword-targeted pages on other domains that don't benefit from the links you earn to your primary domain only creates more work. 100 links to Domain A ≠ 100 links to Domain B + 1 link to Domain A (from Domain B): earning lots of links to "Page G" on a separate domain is far less valuable than earning those same links to a page on the primary domain. For this reason, even if you interlink all of the microsites or multiple domains that you build, the value still won't be close to what you could get from those links if they pointed directly to the primary domain. A large, authoritative domain can host a huge variety of content: niche websites frequently limit the variety of their discourse and content matter, whereas broader sites can target a wider range of foci. This is valuable not just for targeting the long tail of search and increasing potential branding and reach, but also for viral content, where a broader focus is much less limiting than a niche focus. Time and energy are better spent on a single property: if you're going to pour your heart and soul into web development, design, usability, user experience, site architecture, SEO, public relations, branding, and so on, you want the biggest bang for your buck. Splitting your attention, time, and resources on multiple domains dilutes that value and doesn't let you build on your past successes on a single domain. Every page on a site benefits from inbound links to the site. The page receiving the link gets most benefit, but other pages also benefit.

Rules for meta descriptions:

Tell the truth: always describe your content honestly. If it is not as "sexy" as you'd like, spice up your content; don't bait and switch on searchers, or they'll have a poor brand association. Keep it succinct: Be wary of character limits - currently Google displays as few as 140 characters, Yahoo! up to 165, and Bing up to 200+ (it'll go to three vertical lines in some cases). Stick with the smallest - Google - and keep those descriptions at 140 characters (including spaces) or less. Write ad-worthy copy: Write with as much sizzle as you can while staying descriptive, as the perfect meta description is like the perfect ad: compelling and informative. Analyze psychology: The motivation for an organic-search click is frequently very different from that of users clicking on paid results. Users clicking on PPC ads may be very directly focused on making a purchase, and people who click on an organic result may be more interested in research or learning about the company. Don't assume that successful PPC ad text will make for a good meta description (or the reverse). Include relevant keywords: it is extremely important to have your keywords in the meta description tag - the boldface that they engines apply can make a big difference in visibility and click-through rate. In addition, if they user's search term is not in the meta description, chances are reduced that the meta description will be used as the description in the SERPs. Don't employ descriptions universally: You shouldn't always write a meta description. Conventional logic may hold that it is usually wiser to write a good meta description yourself to maximize your chances of it being used in the SERPs, rather than let the engines build one out of your page content; however, this isn't always the case. If the page is targeting one to three heavily searched terms/phrases, go with a meta description that hits those users performing that search. However, if you're targeting longer-tail traffic with hundreds of articles or blog entries or even a huge product catalog, it can sometimes be wiser to let the engines themselves extract the relevant text. The reason is simple: when engines pull, they always display the keywords (and surrounding phrases) that they user searched for. If you try to force a meta description, you can detract from the relevance that they engines make naturally. In some cases, they'll overrule your meta description anyway, but because you can't consistently rely on this behavior, opting out of meta descriptions is OK (and for massive sites, it can save hundreds or thousands of man-hours). Because the meta description isn't a ranking signal, it is a second-order activity at any rate.

Two elements you can control to give the engines context for images:

The filename: search engines look at the image filename to see whether it provides any clues to the content of the image. Don't name your image example.com/img4137a-b12.jpg, as it tells the search engine nothing at all about the image, and you are passing up the opportunity to include keyword-rich text. ex: naming a picture of Abe Lincoln "abe-lincoln.jpg" and/or have the src URL string contain it "example.com/abe-lincoln/portrait.jpg" The alt attribute text: image tags in HTML permit you to specify the alt attribute. This is a place where you can provide more information about what is in the image, and again where you can use your targeted keywords. Here is an example for the picture of Abe Lincoln: <img alt="Abe Lincoln photo" src="http://example.com/abe-lincoln.jpg" /> Use the quotes if you have spaces in the text string of the alt content! Sites that have invalid <img> tags frequently lump a few words without quotes into the <img> tag, intended for the alt content - but with no quotes, all terms after the first word will be lost.

When should you consider a microsite?

When you own a specific keyword search query domain: for example, if you own usedtoyotatrucks.com, you might do very well to pull in search traffic for the specific term used toyota trucks with a microsite. When you plan to sell the domains: it is very hard to sell a folder or even a subdomain, so this strategy is understandable if you're planning to churn the domains in the secondhand market. If you're a major brand building a "secret" or buzzworthy microsite: in this case, it can be useful to use a separate domain (however, you should 301-redirect the pages of that domain back to your main site after the campaign is over so that the link authority continues to provide long-term benefit - just as the mindshare and branding do in the offline world).