Google Sitelinks is an automated program that displays between three and eight indented links for the top ranking website shown for particular search query.  This blog post is dedicated to my research of the Google Sitelinks program and will include a number of examples, theories and best practices.

Preface
On Tuesday October 7, 2008, I spoke at the Search Marketing Expo East (“SMX East”) event at the Javits Center in New York City.  The panel I was part of discussed Enhanced Listings, and my topic in particular was the Google Sitelinks program.

Barry Schwartz of Search Engine Roundtable live blogged the session, and his coverage of the panel is available here.

Finally, as a disclaimer of sorts – I’d like to address the naming of Project Big Water. On the morning of October 7th, my research into the Google Sitelinks program was dubbed “Project Big Water” by industry friends who also shared excellent examples with me to research. Since that day, BigWatah.com has launched, and, Brent D. Payne blogged on Marketing Pilgrim about Big Watah and this effectively unrelated event here.

And with that housekeeping out of the way, lets talk Google Sitelinks.

Google Sitelinks: An Introduction

Google Sitelinks.  What are they? They’re a series of links that Google will display underneathth the first organic search result. You’re probably very familiar with Sitelinks visually – so lets start with Google’s example from the Webmaster Central help center.

Again, this visual is probably quite familiar to you.  But before we begin to dissect these links, lets take a closer look at why Google displays these links to users.

Why does Google offer Sitelinks?

It is my belief that Google chooses to display Sitelinks for a few reasons:

Help users navigate your site: I’ll start with this bullet point because it comes straight out of the  Webmasters/Site Owners Help section. It’s also quite sensible for Google, as their primary concern should be a high quality search experience. Sitelinks will help you to find what you’re after more quickly and easily by eliminating clicks.

To help refine search queries: I’ll venture out on a limb here and say that most people searching Google are lazy. So lazy in fact, that we’ll jump at an opportunity to type less than we already do, even if it means we don’t find what we’re after right away.  Enter Sitelinks. A simple branded query for something like “disney” will allow a searcher to simply click their way onto popular topics and site sections like vacations, Mickey Mouse, or perhaps the Disney Store. Google is interpreting your search and producing immediate refinements based on what they believe to be most useful to you.Here’s an example that I used in my panel presentation for a Google search result on “honda“:

I think the above example is great because it shows us how a brand name like Honda can be interpreted in so many ways. Most consumers will probably be after Honda’s cars, trucks and SUVs — so it’s no surprise to see Automobiles as the number one link.  It’s not unrealistic though to think someone may just lazily search for “Honda” when wanting to check out something about their racing teams, off road vehicles, investor relations, etc.

To link to important pages: Compounding on the previous bullet point, you will often see Sitelinks for the most important pages of a web site.  Why? In addition to being popular resources, important pages are useful and help to improve the search experience.  On smaller web sites you will often find Sitelinks that line up with conventional pages such as the “about us”, “contact information”, “driving directions”, “store locator”, etc.

When do Google Sitelinks Display?

Everyone wants to know how to get Sitelinks to appear for their site. But first, we must look into when they will appear… And that question of “when?” can easily be answered… When one site dominates a search query. We’re talking about sites that are so far in first place, second place couldn’t touch them with all the SEO and backlinks in the world. Unless it was me consulting for them.

In other words, its entirely possible to have Sitelinks display for queries that are not done for brand names.  Since the Sitelinks program debuted in the summer of 2006, more and more non-branded SERPS contain Sitelinks.  All indications are that Google will continue this trend as they look to improve the quality of their search results.

My favorite SERP to demonstrate this is for “video games” where GameSpot is simply killing it with each of the major consoles itemized.  Someone in their SEO department deserves a vacation.

Other times you may see Sitelinks in the SERPs may include important people’s names and proprietary phrases. As an aside, I found it cool that Matt Cutts doesn’t get any Sitelink love but Jonathan Hochman and Stephan Spencer both do.

[Puts on Tin Foil Hat...]

You know what else I think contributes to Sitelinks? Analytical data. Not exclusively Google Analytics, either. I’m talking Google Toolbar data. General usage statistics. Common clickthrough paths. All of those things.

If Google knows that 3,000 people a day are conducting the same search query, clicking through to the same pages in the same sequential order all before exiting a site… Wouldn’t it make sense for them to include a Sitelink to the final destination page?  Of course it would.  And Google has more than enough data to put those paths together.

Theoretical? Of course. But certainly possible.

Fundamentals of Google Sitelinks

Okay, its time to take a deeper dive into the elements at play when Sitelinks are determined and displayed.

The first and most annoying is the fact about the Sitelinks program is that it is an entirely automated system.  You can’t enable Sitelinks.  They’ll either be there, or they won’t be. All of my research has shown that there is a search volume threshold that helps trigger when Sitelinks display.

We need to realize that Sitelinks uses an entirely independent algorithm than the traditional organic search results.  Early indications were that order of Sitelinks was determined by the internal page strength of the pages included in the program. While I certainly believe that internal page strength is a factor involved, it’s certainly not alone.

Another quirk about Sitelinks is that they’ll always display in the same order. You have no editorial control to shift them around at all.

If you want to have any influence at all on your Sitelinks, you’ll want to get your site verified and included into your Google Webmaster Tools account. You may not be able to shift the ordering around or trigger when Sitelinks will or will not display – but you will have some tuning available as I’ll cover later.

The last piece involved here is your own creativity and ingenuity. You can easily drive yourself crazy trying to determine why your Sitelinks behave the way they do… But take it all with a grain of salt.  This is still a very new program with regards to optimization techniques – and it will take a lot of testing to get it right.

Google Sitelinks and Google Webmaster Central

As mentioned above, you’ll need to get your site verified with Google’s Webmaster Central if you hope to influence the Sitelinks displayed for any domain. Once you have done that, you can access your Sitelinks panel through the following navigation:

Once you’re inside that area, things get a little funny on you.  I’ll cover some of the nuances though in the next section…

Frustrations & Quirks with Google Sitelinks

Controlling and Blocking Sitelinks: With the Sitelinks program, there’s no inclusion model Google simply spiders your content, runs your site through an algorithm and pushes out a list of Sitelinks if they deem you worthy. So in that aspect, there is already very little control.

It’s possible though that the Sitelinks Google suggests are ones that you do not want displayed.  For this purpose, Google does allow you to block Sitelinks from appearing in the search results.  I’ve come across a number of cases where this was particulary useful, including here on my own blog.  Many times there is simply content you do not want promoted with the perceived level of authority that a Sitelink carries.

So.. Blocking Sitelinks is easy, right?  Sure it is. Just click on the “Block” link and as Google says, that link will be blocked from appearing for 90 days.

Once you’ve blocked URLs, the blocking period will reset to another 90 days every time you visit the Sitelinks page in your Webmaster Tools account.

No joke. that proves to be problematic when you’re running tests in an agency style environment because there’s no way to check in on one site, see what’s blocked and then leave. There needs to be more editorial control here and I think it’s a very weak assumption for Google to just increase the blocks to 90 days again and again.  You can easily choose to unblock URLs too with just a click inside the Sitelinks panel.

Time frame for Blocks to Stop Displaying: Also, the blocks are not instantaneous.  Test cases that I have run show that there’s an average of about 12 to 14 days before blocks are reflected in the search results.  I’d assume that time frame will shorten up as more people begin using the Sitelinks program… But for now, it is what it is.

Three to Eight Sitelinks: That’s the total number of links that will display in the search results. If you have an active inventory of more than 8 Sitelinks appearing in Google’s Webmaster Tools – start blocking out as needed to make sure the top 8 there are the ones you really want to be used.

You can always block out more, too.  You don’t need to fill up all 8 slots. You do need to maintain at least three unblocked Sitelink URLs though.  Any less than that and the system will stop displaying your Sitelinks alltogether.

Case Sensitivity: Another quirk with Sitelinks is their inability to maintain character case in links.  Everyone is familiar with Apple and their iPhones, iPods, ITunes and so on… But check out how Google displays their Sitelinks:

Getting New Content into Sitelinks: Ready to be discouraged? For the sites inside of my testing group, the newest content showing up in the Sitelinks was more than 7 weeks old.  That’s horrifying for sites (including blogs, in particular) that thrive on current events and stories.

Google Sitelinks: Best Practices

Sitelinks is simply an enhanced type of search result.  Therefore, the standard SEO variables are in play here, with a heavy emphasis on links – both internal and external links.

You’ll want to make sure you structure your content in a typical manner. Stick to content themes and employ the SEO101 approach.  You’ll want to use headings where appropriate to organize content on pages and pay close attention to providing clear and concise page titles.

Since link popularity and page strength are both important factors in the Sitelinks algorithm, be sure to measure links using the Webmaster Central tools. As is the case with most Google SEO, the more links – the better.  Just be careful not to rely too heavily on any one type of link or anchor text.  Subtle variations seem to help Google understand the importance of Sitelink candidates.

If you care about your Sitelinks – never 301 and existing Sitelink.  In three seperate case studies that I ran earlier this summer – all three sites lost the Sitelink URLs (old, and new).  One of those sites had been showing 7 Sitelinks, and one 301 somehow shut down all of the Sitelinks in a few weeks time with no other changes having been made.

Google Sitelinks & XML Sitemaps

Kate Morris detailed a test that she had run where the priority supplied in an XML sitemap actually influenced the ordering of pages in the Sitelinks program.

Like Matt McGee, I’ve never found XML Sitemaps to be much of a help at all.  But, I can understand how (or at least why) Google would be open to considering this data for shifting Sitelinks around.

I haven’t had a chance to test this out myself yet – but I do have a test planned for the next few weeks.

Using Social Media Tagging to Score Google Sitelinks

One of the posts here on my blog that I needed to block was a blog post about someone.  That post had a number of people tagged too.  In less than a month that entry shot up to the top of my Sitelinks which caused some concern.

It’s not that it was bad content – it’s just not that important to users who are looking for me by name.

In researching the backlinks on that particular post, I found that Technorati was carrying major influence since many of the backlinks were reported on www.technorati.com in addition to the various sub-domains of technorati.com including feed.technorati.com, search.technorati.com and others.

Other social media sites that used tagging to help create a massive amount of backlinks include Friendfeed, Twitter and MyBlogLog.

Questions? Comments?

Admittedly, the topic of Google Sitelinks can be a bit dry and plain.  When used correctly though, Sitelinks can really improve upon Google’s user experience while also providing more opportunities for search marketers.

But – this is still a very new program in the sense that Webmaster Central control is only a few months old. If you have any questions on this program, I’ll certainly do my best to help out. Leave a comment below so we can help to grow this resource out for more users to benefit from.

It’s a long ride from our office in Coventry Rhode Island but we’re putting our SEMNE membership to use! I’m joining three other team members and two designers (including Hunter Satterwhite) from the ADP and BZ Results teams, and we’re making the 99 mile drive to Trumbull Connecticut for tonight’s Search Engine Marketing New England (SEMNE) event,

The SEMNE event was originally slated to have Vanessa Fox‘s panel of Using Marketing to Enhance Your Search Traffic. Vanessa will be unable to make it due to the passing of a family member though, and our thoughts are with Vanessa during what’s certainly a difficult time.

Tonight’s presentation of Using Marketing to Enhance Your Search Traffic will discuss how marketing can (and should) be used to enhance your search engine traffic. With speakers Chris Elwell, Dawn Briggs and Robert Jackson – I’m really looking forward to tonight’s event.

If you’re in the NY Metro Area or Connecticut and can make the drive, I’d love to meet up and talk search with you. Here’s the complete event details…

Trumbull Marriott Merritt Parkway
180 Hawley Lane
Trumbull, CT 06611
(203) 378-1400
Get directions

6:30PM-9:00PM

Speakers

• Chris Elwell of Third Door Media
• Dawn Briggs of Business.com
• Robert Jackson of Hochman Consultants

Admission
Admission is free to SEMNE Members and $49 for non-members.

Online Registration
Online registration is available for this event.

Thanks to the SEMNE Crew
As always, much thanks goes out to Pauline Kerbici and Jill Whalen, the co-founders of SEMNE – as well as Jonathan Hochman who helps with the CT based events and coordination.

The Last SEMNE Event: Getting Into Google
I haven’t had a chance to make it to many of the events since we bought our corporate membership, but I have found them to be very rewarding on a number of levels.

The last event I attended was back in May of 2007 and it featured Dan Crow, Director of Crawl Systems for Google. SEMNE has a complete write up of that event including a photo of Brandy up there. Jill also had a write up in High Rankings that’s worth a read.

I’ll try to get a full post recapping tonight’s event on the blog in the next day or so.

Don’t froget to find me if you’re going!

It’s no secret that Digg, Reddit and other news and media aggregators are a target for search marketers. In my quest to learn more about the intricacies of these social environments, I sought after a community to review and understand a bit more.

The trouble I ran into was that with sites like Digg was that there are too many active users and too much going on at any one moment to form hypotheses on.

While analyzing the marketing opportunities of a socially active environment may seem difficult, the one I ended up reviewing was quite simple. Since the title of my post eliminates any element of suspense, I’ll plainly say Sphinn was the community I opted to dissect.

What I found was some great potential, but a number of obvious problems that are preventing the community from growing.

** Please note, this post was a draft for about 3 weeks, so now that it’s live, the data is a few weeks old. **

Background / History
First, let’s understand the purpose of Sphinn. Created by Search Engine Land and Danny Sullivan, Sphinn was officially launched on July 12, 2007 for the purpose of aggregating newsworthy items in the search industry. The links featured on Sphinn could come from blog posts, news portals, forums, etc. At the end of the day though, content was said to be specific to the search industry.

Basic Data on Community Users
For me, a community is only as rich as the number of active participants. When I pulled my data there were 12,410 registered members according to the alphabetical listing of users shown in the Sphinn Network.

Now, I watch Sphinn on a regular basis. I’m positive that 12,410 is not a fair representation of the active user base. In fact, I’d say that number of really active members is more along the line of 124, or just 1 percent.

Continuing to pull numbers using the available search features, I came up with a few key stats:

Registered Users: 12,410
Users with 5+ Submissions: 750
Users with 2+ Comments: ~1,300
Users with 1+ Sphinn: 6,259

My Initial Reaction…
Are you flipping kidding me?

Communities, like all social groups tend to have two types of users – leaders and followers. With Sphinn, I’d consider the leaders as those making the most submissions and comments while the followers are the pile jumpers who just vote for things to see their own stature improved upon.

Now – as Gerhart would surely tell me, this is a community where stardusters look to promote their own work. When they post, they hit up friends and colleagues to Sphinn up their stories to help cement their legitimacy in a rather veiled, niche industry. (Of course, I’m okay with that, especially because people often ask for Sphinns and I tend to reciprocate that behavior because it’s free traffic.)

The Leaders (Submitters and Commenters)
750 members, or 6.04% of the community has more than five submissions. Five. While I love to set really low expectations so they can easily be met, even this surprised me.

Unless I’m entirely off base here, and I know I could be – I believe that five submissions would classify a casual user of the network, not a leader. Here’s a graphic I made because Excel likes to do that for me:

Looking into it further, I thought okay… My standards are too high. maybe the leaders don’t really submit things, but contribute on a regular basis in their commentary of submitted stories.

It took lowering my expectations to get the numbers to improve. There were about 1,300 users who have submitted multiple comments on Sphinn. That’s less than 10.5%. For those who like graphs with complimentary colors, I give you this:

Hyperactive Users
Or maybe I mean hyperactive ABUSERS? You be the judge.

First, my apologies to Bill Hartzer and Steven Bradley. I’m using your activity on Sphinn to prove a couple of points here, and in exchange for that I’m giving you each a backlink without no-follows for good measure. I have never (knowingly) met either Bill or Steven, so I apologize if this offends anyone. I’m not trying to judge any characters here, just user data as provided to me on Sphinn.

Bill Hartzer, The Sphinnmaster
Bill Hartzer (bhartzer) has cast 10,427 Sphinns. The next active user, Maki (DoshDosh) has 5,108 Sphinns. That’s 48.9% of the actvity of Bill. In other words, Bill’s out there giving our Sphinns like candy on Halloween. Stopping short of accusing Bill of just sphinning everything he sees, I’ll say this… Bill has 1 comment for every 100 sphinns cast. Continuing the comparison, DoshDosh has nearly 5 comments per 100 sphinns cast. aimclear, Sphinn’s third most active user produces nearly 12 commens per 100 sphinns.

I’m sorry Bill, I just don’t see how you can vote for that many items without providing more commentary. I’m certainly open to your response though. Am I just assuming too much on the numbers?

Steven Bradley, Speed Sphinn’er
Steven goes by the name of VanGogh on Sphinn. He sphinns a lot of Search Engine Journal’s posts. That pleases me. While wrapping up this post though I clicked on over to Sphinn Live. Here’s what I saw (my notes in red):

In a 16 second span, Steven read and voted up three stories. Then, Bradley picks up by giving out another 2 votes in a 44 second time span.

If Stephen’s sphinns are truly a vote of authority or approval, then Steven read 5 articles in 66 seconds, and judged their contents highly enough to vote for them on Sphinn.

(added hours after initial post)
The Community Responds!
Please Note — Since the time this was posted, I’ve gotten some great feedback and responses on my thoughts. Particularly, I want to applaud Bill Hartzer’s blog post and Steven Bradley’s comment. Both of these gentlemen took the position of absolute professionalism – and their thoughts and counterpoints need to be read.

Furthermore, Danny Sullivan offered his thoughts on my post as well on the Sphinn story page for this post. Please be sure to read the comments there in conjunction with this blog post.
(end of post-launch additions)

Don’t Bite the Hand That Feeds!
I’m not bashing Sphinn directly here, so please understand that. The only thing I’d like to do here is raise some concern for what has quickly become a reputable site and resource for those in the industry.

Sphinn provides a decent amount of traffic to Search Engine Journal and it has done the same for my blog here too. However, I’m afraid that some activities need to be policed a bit more.

Should a user be able to vote up 5 stories per minute?

Should users be permitted to continuously vote for things without contributing more comments, submissions or feedback to the community?

Should user accounts be archived or removed after a specific duration of inactivity?

Is there Enough Data?
Rather than come off as accusatory or assumptive, I do understand that there could be too small of a data sample to work from here. What do you think? Is there enough data here to warrant making changes to help the long term growth of the Sphinn community? Can the community continue to grow beyond the limited reach it has already required?

---

These are only a few questions that come to mind for me, and I’d love for these thoughts to be heard by the Sphinn Staff. That’s all I’ve got for tonight, but I’m all ears when it comes to readers comments. Of course, I’ll be submitting this to Sphinn for good measure… Gotta cash in on that free traffic!

We’re paying top dollar to be told that we’re either bullshit artists, or wasting away our time and careers. That my friends, is bullshit.

The search marketing industry is all frenzied now that Jason Calacanis decided to upset some of the best SEOs in the industry. Again. The question I’m left to ask is — Why are we really surprised?

When Calanis’ stated that SEO is bullshit in his keynote at SES Chicago, everyone was ready to tear him a new one. Neil stepped in, proved him wrong, and then slowly, Calacanis repositioned himself as someone the industry paid attention to.

Yesterday in his keynote at SMX Social Media — Calacanis infuriated the industry again, prompting some prominent figures to literally walk out. This time, we’re being told that SEO is a wasted industry, and that we’re fighting off ranking problems instead of creating good content.

Unable to say anything without reaching around to pat his own back, Jason then adds that using a human service is a better way to go about it. Really Jason? I bet Mahalo would qualify as a human powered service, right? Sigh.

As an industry, let’s just put ourselves in a position to stop dealing with this type of juvenile behavior. How you ask? With one simple step. Get him off of the conference keynotes.

The truth is, SEO’s don’t care about Mahalo. It’s a no name search portal that uses biased human editing to control SRPs. Notice, that’s not a SERP — because Mahalo is not an engine.

Furthermore, Calacanis knows next to nothing about SEO. While I will say that he’s more related to SMO than he is SEO and SEM — people who are paying a premium to attend events like SES and SMX are really tired of his whole charade.

Jason Calacanis is quickly becoming that annoying ass kid in the restaurant that screams bloody murder when he doesn’t get his own way. He’s doing this whole thing for attention — and all we’re doing is setting him up to do it again and again and again.

This blog post shouldn’t even need to be written, because like Calacanis — it’s irrelevant to the search industry. I just hope some of three three readers I have here on my blog will hear me out and turn away from clown.