Google Search Engine Optimization

January 27, 2010 · Filed Under google, seo · Comment 

On January 21st, 2010 the Google Webmaster Central Blog released a report on Google Search Engine Optimization (SEO).

They’ve also released a video that was recorded in May of 2009 at the Google I/O conference (the same conference that first introduced Google Wave).

Now, the video below is an hour long so if you feel up to it and have the time, feel free to watch it yourself.

If not however, you’ll be happy to know that I in face HAVE watched the video and have made the important notes about Google Search Engine Optimization below.

The focus of the presentation was based on organic search or natural search results.

“It’s not enough to make a site and stick on the web because there’s so much out there.  You actually need to make it crawlable so that it appears within the search results [of Google].  And there’s steps that you can take along the way to be sure that your site appears within the search results.”

If you take a look at your site within Google’s Webmaster Tools area you might notice that the total URLs that Google found is larger than what the Indexed URLs are within your site.

“We can’t index everything because a lot of it’s boilerplate duplicate content information.”

It’s important to verify your website using Google’s Webmaster Tools.

Planning for a New Site

  • A compelling product or service.
  • I have unique content that I want to create.

“If you have a great service or you have great content you want to look to make a site.”

Selecting Your Domain Name

You want to look at what you want your site to target.

To target international vendors from all over the world, look to get a .com, a .net or a .info. (gTLDs)

If you want to get a ccTLD (country specific domains) then you can only target people in that region (at least through Google.)  For example, if you get a domain such as www.SomeSite.fr then it’s only going to target customers in France not in the United States.

“If you want to get the most authority, then you’ll want to self-host [your site].”

It’s important to choose a non-www or a www version of your site because of PageRank dilution.  It will disperse your links between two sites.

If you want to target users who only speak a certain language then you don’t want to use the Geographic target tool in Google Webmaster Tools.

Determine Where to Host

If you don’t give Google an indication of who you are serving (country specific domain or Geographic location setting in Google Webmaster tools), then you’ll want to choose a webhost that is in the same location as the location you’re trying to target.

“Shared hosting with “bad neighborhoods” does NOT affect your rankings.”

There’s new Google Webmaster Tools API data available.

Designing Your Site

Site Architecture:  Navigation

  • Important that users are able to navigate your site.  For example, breadcrumbs.
  • A Search Box
  • Logical category navigation
  • If they come to a child page, how can they get back to that home page?
  • Know that they’ve already visited.
  • Modify color for visited links or “Items you also looked at”
  • Site Architecture for Users:  URL Structure

“Shareable URLs are always good.”

  • The URL takes a user to the exact item they clicked on, not a menu or a homepage where they would have to search for the item.
  • “Descriptive file names also help users as well as search engines.”
  • Keywords separated by hyphens.
  • Lowercase URLs for “safest implementation”.  For example, users might manually type it in and they’ll use lowercase.
  • Keep in mind that robots.txt is case sensitive so mixed-case URLs can be a “hassle”.

Site Architecture for Search Engines

First think about what’s going to be private vs. public and secure the private data.

If you have a site that has adult content – separate the adult content from family-friendly content.  Especially for safe search.  It’s easier for the search engines if they see a directory structure that’s logical.

For example:

www.SomeSite.com/adult-images/

www.SomeSite.com/family-friendly-images/

“For search engines, static links are always best.  Because we can follow those and they have good anchor text.”

  • Keep important pages well-linked from the home page.  Don’t hide it in a child node, link it from the home page.
  • Submit an XML sitemap with Yahoo!, Ask, and Microsoft as well.  Often these are from sitemaps.org.

Disallow Throwaway Content:  Shopping Carts

Since Googlebot will never actually “purchase” a product from your site, there’s no need for shopping cart code.

Disallow Login pages as well.

Create and test the Robots.txt file within Webmaster Tools.

Dynamic URLs

Often recognized by name/value pairs. Eg:  page.foo?category=1&answer=10

Create easily understood name/value pairs:

http://www.SomeSite.com/product.php?item=g1-phone

http://SomeSite.com/product.php?item=g1-phone&category=mobile

http://www.SomeSite.com/product.php?item=g1-phone&category=mobile&affiliateid=1234

http://www.SomeSite.com/product.php?item=g1-phone&category=mobile&affiliateid=5678

Each of these lead to the one product page for the g1 phone.  But they differ from one another because of all of the different parameters.

You might have 1,000 products on your site and you might have 10 URLs that could actually link to each product so now you’ve created a site with 10,000 URLs and actually you only have 1,000 good product pages.

Google can’t crawl all the web.  ”We’re trying to limit our resources and crawl what’s useful.”

Since each category is the same, and the affiliate ID’s differ only slightly they’re able to throw those out as not being a single, different URL and makes it much easier for Google to process.

They’re going to find out ways to crawl the most important of those [URLs].

“One reason why people might not have done this in the past is they thought “ranking reasons”…they thought “oh, I need to put keywords in my path because that’s better for ranking.  Which isn’t the case.”  We’ll actually interpret keywords whether they’re in the page or in the name/value pair.”

Remove Session IDs from Paths or Position

Creates undesirable infinite crawl possibilities.

If you put your session ID in the middle of a path, it looks infinite to Google.

“Our search engine wants to crawl every category and every leaf node page. ”

If you have a lot of navigation, expand one category at a time.  Keep it standard.

Google’s Message Center in Google Webmaster Tools

They’ll give you a notification if they’re found an infinite URLs.

Indexing and URL Structure

  • “You want to have pretty sweet anchor text.”
  • Create descriptive anchor text that names or labels your link.
  • You don’t want “click here”.

Choosing Technologies

Design initially with static HTML content and navigation to reach the largest audience.

Add Flash & AJAX later.

JavaScript

Google now processes many onclick events to discover new URLs.

This functionality is continuously updated.

Can also call an onclick if the onclick calls a function but ONLY if the function is on the page itself and not dependent on external files.

Examples of what Google can now process on your site:

div onclick=”document.location.href=’http://somesite.com/’”

if myfunction(x) is defined within the page:

onclick=”myfunction(’index.html’)” <a href=”#” onclick=”myfunction()”> New Page</a>

<a href=”javascript:void(0) onclick=”window.open(’welcome.html’)”> open new window</a>

They’re not yet there with AJAX.  Consider Hijax for AJAX.

For Flash

Google can find all of the user-visible text in a Flash file.

Avoid text in graphic files if you desire content indexed

They associate text with the embedding page (if one exists)

For embedded Flash, use well-known JavaScript methods to do the embedding:

  • SWFObject
  • SWFObject2
  • UFO
  • AC_RunActiveContent

New functionality added by Google:

If your Flash file loads an external resource (e.g. HTML page, XML file, another SWF file), they’ll capture that content and associate it with the Flash file.

Frames & iFrames Current Behavior with Google

Crawl

  • Frame/iframe URLs discovered at crawl time within the parent page.
  • URLs do not require a separate link on the web

Index

  • Frames often indexed with parent page.
  • iFrames often indexed as an independent URL.

Historically we’ve seen a lot of ads within iFrames and so there was no use to bring that into the parent page.

Making the Web Page

Anatomy of a search result.

samplesearchresult

There’s a title which corresponds to the title of your page.

A Snippet  which can either be found within the context of your page or used from your meta description.

URL and sitelinks.

Sitelinks are calculated algorithmically done especially if you have a site that’s very accessible to search engines (and it’s going to be the high result.)

You want to create unique titles for EACH URL.

As well as a unique meta descriptions.

Use rel=”canonical” to specify one URL from others with duplicate (or near duplicate content).

Minimize Duplicate Content

If you have multiple URLs that go to the same page this means you could be losing PageRank.

Specify your preferred version in the header of each duplicate page (Yahoo and Microsoft accept this standard).

For example:

<link rel=”canonical” href=”http://SomeSite.com/product.php?item=g1phone” />

Properties such as link popularity will be consolidated to one version.  The canonical URL will also likely appear in search results.

Specifying canonicals only work within the same domain.

Google Webmaster tools can help with titles and meta descriptions.  It will point out to you if you have duplicate titles or duplicate meta descriptions or titles that don’t exist.

Adding Content

Make sure you’re keywords are included in the text.

“Keywords are important.  If you want to rank for a certain word, you should use it in your text naturally.”

Adopt the language of your users.

“Keyword stuffing is spam.”

Images

For image search here are some of the signals:

  • Google looks at the text surrounding the image.  The paragraph text and the heading.
  • The quality of the image matters.  The higher the better.
  • Image attributes are very important.  Often, the alt text is used as the images anchor text.
  • Image file names.  Use file names that are hyphen-separated rather than a default.
  • “Hyphens will be seen as word separators whereas underscores are not seen as a full-strength word separator.”

Image Replacement:  Hiding Text Can Be Risky

Developers (or some site owners) will often think that “Oh, I’ve got this image with some text in it so to let search engines know about my text, I will actually use “text-indent:-999″.

What this does is push the text off the screen so no one can see it.

This is a bad idea because this is a long-running spam technique and will generate a penalty.

If you’re using images, use alt text if possible.

If Google believes that you are a legitimate site trying to use something like the text-indent, they’ll send you a message in Google Webmaster Tools.

For Videos

“YouTube does a lot of good things.  They have static HTML navigation, text descriptions, good titles, you can visibly see the popularity of a video so you can see the views, the ratings, they have comments that help the user become engaged; comments can also be a turnoff but it also keeps some people there.  They have related videos that keep it sticky on the site so you keep looking at more…and of course the best thing about video optimization is to create a compelling video, right?  One that people do want to share.  And if you’re hosting on your own, uptime is important.”

Video Site Maps

You can submit a video sitemap to Google.

Responding Appropriately (Response Codes)

200 – For when the URL exists; No “soft 404s” that redirect to your homepage.

301:  Signals the content has moved permanently; Use when existing content has new URL; Transfers URL and linking properties from source to target URL.  Search results will show the target URL not the source URL.

404:  File not Found; Google treats 410 similarly

Webmaster Tools allows you to copy and paste code that will give a nice looking 404 with the closest match to the URL they were trying to get to.

503: For temporary site maintenance, signals to crawler to revisit later.  Don’t noindex everything and don’t 404 it.

Review Crawl Errors

Google Webmaster Tools will give you response codes for errors.

Fix Broken Links!

Webmaster Tools provides the source (within your site or external to your site) for 404s.

They’ll tell you where that 404 is from.  Can reach out to people linking to you who are coming to a 404 and let them know that they’ve mistyped something, etc.

It is up to you as the site owner to attempt to correct this.  You can redirect the page they’re linking to using a 404.

301s don’t tend to get visited and crawled quickly.

Sitemaps

There’s lower-case sitemaps.  The HTML type of sitemaps.  SE’s can follow that but what Google likes and wants is an XML sitemap.

You can put the link to the XML sitemap in the robots.txt file.  This is a list of URLs that you WANT crawled.  They’ll use this to feed the URL to their crawler.  It maximizes the pages being crawled by Google.

“It’s best to have strong link architecture in your site.”

Engaging the Community

Blogs

  • Become an authority in your field.  Don’t just “put something out there.”
  • Use categories that are also good keywords.
  • Keep your blog secure; Hackable software warnings in Webmaster Tools Message Center
  • Review http://www.MattCutts.com/blog/whitehat-seo-tips-for-bloggers/; or just Google [matt cutts seo blogging tips]

P.S.  On this site there is a link to a video where Matt Cutts talks about White Hat SEO Tips for Bloggers from Wordcamp in 2009.  And yep, I’ve added the finer points to that 45 minute video as well.

Blog FAQs

  • Publish full content or a snippet?  Only mildly better for Google if full content is published.  Your preference.  ”Most users prefer to see the full content.”
  • Robots.txt disallow duplicate content categories/labels like “archives”?; Sure, if you’d like Google to only crawl unique sections that’s just fine.
  • What about tweets?; Not currently in ping service, crawled if linked to on the web, can appear in web search results.
  • Twitter has no ping service.  It has to be linked to on the web.

Engaging the Community:  Comments/Reviews

Benefits

  • Informed purchasing decision for consumers
  • Communication with/between readers.

Host comments and reviews on your own site for stickiness (need to moderate for spam).

Place reviews on same URL as product to consolidate content (all related links go to the same URL).  Keep the reviews on the same page as your product so that they don’t go to a separate URL to read a review.

Microformats and RDFa Reviews in Snippets

Initial focus is on sites whose primary intent is to provide

  • Reviews
  • Descriptions of people (e.g. social networking sites)

droolingdogreveiw

If they’re marked up in RDFa or in an hCard we can actually show that on Google.

Check Backlinks in Webmaster Tools

Find out who is linking to you in Google Webmaster Tools.

Maximize Traffic

Google shows you up to 100 queries now of how people are finding your site.

Final Question to end the session:

“Do you think it’s better to create a good website or to start a social media campaign?”

Google’s answer:  A good website.

“A lot of people make conversions when they feel they’re well-informed and comfortable.”

Tags:

Next Page »