- Landing Pages: It’s impossible to do a good job of optimizing your homepage for every possible term people might use to find your site. Think of it as a town fair full of criers who are all yelling their own messages: the end result is a din of roughly equal volume in which nothing stands out. Plan instead to add a page to your site for each search term, heavily optimized for that term using all the tips below, so that page becomes the top organic search result for the term and therefore the page that visitors land on when coming to your site. It’s important to make sure that these pages aren’t islands (i.e.: not linked from any of the site’s main content), because otherwise web crawlers may not find and index them.
- Titles: Some of the most overlooked SEO real estate in the world is staring right at you from the top of this very window. The
<title>tag, which sets the text displayed in the title bar of the browser window, is very highly rate by search engines as being indicative of the page’s content. The engines differ in how much of the<title>they index, but the general rule of thumb is that the first 60 or so characters are the most important. This dictates that the search term should come before things like a company name, so it would be better to have “8 SEO Tips and Tricks » JayGoldman.com” rather than “JayGoldman.com: 8 SEO Tips and Tricks”. Luckily, this also tends to be more useful to users when they view their browser history or bookmarks in a narrow window or menu that cuts off the text, since the name of the page they want is more likely to be visible. I use the WordPress SEO Title Tag Plugin to swap the order around on this blog. - Repetition: The search term should be repeated in an
<h1>as close to the top of the<body>as possible. We saw a difference for some of Radiant Core’s clients between having text at the top of the HTML and moving it down for presentation using CSS and just putting it at the bottom (e.g.: the list of SEO links at the bottom of the TargetVacations site actually occurs at the top of the HTML and is moved down through a combination of CSS and JavaScript since the page’s length is variable). The term should be repeated again in a<p>following that<h1>, ideally surrounded by<strong>tags. - Font Replacement: A necessity if you’re particular to a specific font and want to make sure your text is rendered in it. Since HTML doesn’t yet support embedding fonts (though it’s coming in CSS3 as WebFonts), specifying a font in CSS will only work if the person viewing your site has that font installed on their computer (and could still look strange if they have a different font with the same name). There are two popular routes: image replacement and sIFR for Flash-based replacement. Image replacement is much more limiting in that it requires you to create an image for each piece of text, while sIFR can be really difficult to get working, requires Flash for display, and can really slow down page rendering. I use a mix of the two on the homepage of this blog, rendering the header using image replacement since it never changes and rendering blog titles in sIFR to get Futura without having to manually create images for each post’s title.
- Doug Bowman of Stopdesign (and now the Visual Design Lead at Google) has a great overview of using background-image to replace text. My preferred method was originally outlined by Mike Rundle and has gone on to be the favourite used widely by web designers (and is even linked to by Doug): Accessible Image Replacement.
- The concept behind sIFR is really elegant: create a very lightweight Flash movie that has the font embedded and then pull in the text it replaces and render it using that font. The accronym stands for Scalable Inman Flash Replacement, named after Shaun Inman who came up with one of the original CSS-based image replacements. sIFR was originally created by Mike Davidson and Mark Wubben but hasn’t been updated by them in a long time. There’s a sIFR Lite available from AllCrunchy.com, though it looks like it hasn’t been updated in a while either.
- Domain Names: Most of the things search engines look for center around trying to determine the content of a page based on the text it contains and the meta information that surrounds it. The more difficult it is to fake the meta data, the more stake placed in it. Some of the hardest to fake are the domain name and URL of a page, which makes them two of the more important tweaks you can make. It’s harder to optimize the domain name since you only have one for your whole site, but if you sell You Won’t Believe it’s Not Tuna you should grab a domain like tunareplacement.com rather than chickenofthesea.com. The age of your domain name does factor into the calculation, so it’s generally better to renovate a site and keep the old domain than it is to start entirely from scratch. It’s also worth noting that some search engines, particularly Google, treat subdomains as different sites, which means things like blog.jaygoldman.com and www.jaygoldman.com don’t necessarily share PageRank. Unless there’s a stronger-than-SEO reason to go with a subdomain, consider a directory instead (www.jaygoldman.com/blog). You should also consider that www.jaygoldman.com and jaygoldman.com (without the www) aren’t necessarily the same, so you should decide early on which you’re going to use and be consistent in promoting the site (I use jaygoldman.com). You can configure mod_rewite (see below) to remove the www if you choose to go that route.
- URL: The search term should ideally be part of the URL, using -s for spaces (e.g.: www.tunareplacement.com/recipes/tuna-and-marshmellow-salad). This is much, much better than the default URL that your blog sofware or CMS probably produces (www.tunareplacement.com/recipe.aspx?id=23), so you should absolutely switch over if you have that control (WordPress users should take a look in the Permalinks section of the Settings in their WP-Admin). I’ve always prefered avoiding file extensions in URLs entirely (e.g.: .jsp, .php, .asp(x), etc.) since it exposes part of the site’s implementation into the URL and then into people’s bookmarks, web crawlers, and the like. You’ll break all of those if you later rebuild the site on a different technology, so it’s better to abstract to a higher level earlier and just change the rewrite destinations later. Human readable URLs also kick machine generated URLs butt when it comes to things like analytics or emailing links to friends. I much prefer using URL rewriting, which allows the clean and human readable /recipes/tuna-and-marshmellow-salad to get rewritten to /recipe.php?title=tuna-and-marshmellow-salad behind the scenes. If you’re running Apache and don’t mind some server config, take a look at mod_rewrite, but be warned that it’s like black magic, ninjas, and awesomeness mixed together in a very potent but incredibly tricky potion. If you’re running IIS, take a look at IASPI ReWrite, IIS Rewrite, or Mod Rewrite for IIS. I’ve got no experience with any of them so that’s about all I can say on that topic.
- Sitemaps: Way back in the early days of the web, Site Maps were actually a page on your site that showed people where all the other pages were, usually in some sort of graphical flow chart fashion. Most sites have grown considerably beyond being representable on a map, but they’ve found a new lease on life thanks to web crawlers. Submitting a Sitemap XML file to the search engines helps them understand how to crawl and index all of the pages, including the frequency that they change. You really don’t want to have to do this manually since it has to be updated every time a new page is added, so take a look at automated tools that will do it and submit the update (I use the Google XML Sitemaps plugin for WordPress for this blog). More info at Sitemaps.org.
- Inbound Links: You want to encourage as many inbound links to your site as possible since they are factored into most search engine’s ranking algorithms as essentially counting as votes for the autoritativeness of your site. Almost all inbound links are positive, with the exception of ones from things like known link farms, but you really want to focus on getting other sites to link to your landing pages with the right link text. If we’re trying to optimize the Recipes page of our Tuna Replacement site for the search term “tuna recipes”, it’s much more valuable for outside sites to link to that page with
<a href="http://www.tunareplacement.com/recipes">tuna recipes</a>as the link than it is for them to link with<a href="http://www.tunareplacement.com/recipes">recipes</a>. If you have the kind of site where people might want to feature your content elsewhere (with, say, a Tuna Recipes Widget of the Day), consider developing an embeddable form that includes links formatted to match your SEO priorities. - Meta Tags: These used to be all the rage in that you could define keywords for search engines to use in their indexing. That’s a pretty easy system to game (want to attract attention to your porn site? Try keywords like “free money”), so they’re no longer nearly as valuable as they used to be. There’s a lot of discussion in the SEO community about how valuable they actually are, but the general conclusion is that you can’t go wrong by adding the keyword and description meta fields to your pages, and that they may even be used to display some of the information in search results. I use the Add-Meta-Tags WordPress Plugin to automatically add them to all of the pages and posts on this site.
Senin, 27 September 2010
8 SEO Tips and Tricks
A client recently asked for a quick overview of good SEO practises, so I thought I’d share them with all of you at the same time. This is by no means an exhaustive list, but following these items will definitely make a difference to your site’s performance in major search engines. These are roughly in order of priority (the first items being the most important):
Langganan:
Posting Komentar (Atom)
Tidak ada komentar:
Posting Komentar