Getting New Sitelinks Under Your Top Listing At Google

[gugo]

How is possible to show the results with subtopics in Google?

I want to show the main web site and the subtopics of each one.

Example here with the "advanced web ranking" keyphrase.




BTW: The MAIN results are explained in the next link by Google, but nothing about the subtopics posted in the image before.

http://www.google.com/help/interpret.html

[SEMBasics]

What makes you think that it is possible to do this? I looked at the link to Google that you included and they didn't mention anything about displaying subtopics. Furthermore, I did a search on the Google network for subtopic and nothing showed up which relates to what you are asking about.

Do you have reason to believe that this is possible and just want to know how to do it, or are you asking is this possible to do on Google?

Furthermore, is this possible on Yahoo and/or MSN? If so, could you tell me how? Thanks.

Best of luck,

Moshe

[gugo]

Precisely i´m asking if the display of the subtopics is possible by:

• an internal structure of the web site created speacially for the Google bot or

• a web site standard (W3C) and only if your are in the 1st position appear the subtopics results.

[gugo]

Other example in generic competitors.

I put the brands and the results are with or without subtopics in the same 1st rank.

[pleeker]

You have no control over this. Google does this algorithmically only on certain brand name searches where the algorithm can be almost certain of your search intentions.

For example, search for [fedex] and you get subtopics -- there's only one FedEx. But search for [ups] and you don't get the subtopics. Why? Because UPS also stands for "uninterruptible power supply," University of Puget Sound, etc.

HTH.

[Robert_Charlton]

I'm glad to see this topic surface again. I put up a long post on a thread about the subject back in April, and got no responses. The subject has been hashed over a lot, but no one has answered why Google chooses the particular links it does. Here's the post...

Additional links in Google results
http://forums.searchenginewatch.com/...9560#post79560

Re gugo's question...

Originally Posted by gugo

I want to show the main web site and the subtopics of each one.

...I don't think you have much control over this, except to have a predominant number of pages on one domain that would rank in the top 10 for your search. It's likely this will happen on a company name, but, as you observe, it doesn't happen with all companies.

Noting your particular example of Ford and General Motors, I think those results are pretty easily explained. Ford is competing with some other Fords, and Ford also uses multiple domains a lot, so there's ford.com, fordvehicles.com, ford.ca, ford.co.nz, etc, as well as the other Fords, like Gerry and the modeling agency.

General Motors, on the other hand, appears to use directories instead of additional domains or subdomains, and there haven't been many army generals named "Motors."

I would love to get some thoughts about my last comment in the above-cited thread, as I think it might shed some light on how Google treats multiple links on a page going to the same destination page...

Originally Posted by Robert_Charlton

Some of the bottom links go to the same pages that the side nav links go to. It's not completely obvious why, when there was such a choice, that Google picked the links it did.

[pleeker]

Very interesting, Robert. I've never really analyzed the choice of links since none of my small biz clients get this red carpet treatment in the SERPs.

But without going to Google, I just thought back to my Fedex example and made a quick mental list of the page links I'd expect to see below the main home page listing. My list was:

- track a package
- locate a drop-off point
- see shipping rates

And then I ran the [fedex] search on G, and none of my guesses is there. Instead they're showing

United States Shipping Home
United States Shipping Service Info
Site Map
Investor Relations

Seems odd, but it looks like it's due to the way FedEx does their root home page ... not much there beyond a geo-dropdown, an About link, and the Investor link.

Hmmm.........

[Chris_D]

Hi guys,

The closest answer to this was matt Cutt's comments when the feature first hit back in August last year.

For a small number of sites, we’re not just showing our regular snippets: we try to expose useful links from within a site. In this Berkeley example, Google shows links for Berkeley departments, academics at Berkeley, etc. Pretty neat (and more importantly, useful) stuff.

People who know Google well will go “Cool” and move on. Other folks will ask things like “Are sites or their links selected by hand–can my site get in on this? Is money involved?” And the answer is: it’s all algorithmic. The algorithms pick the sites where this could be helpful. Of course money isn’t involved at all.

http://www.mattcutts.com/blog/ui-fun-better-snippets/

I interpreted Matt's comments - "its all algorithmic" and "expose useful links" as indicating the links selected to be displayed are most likely going to be the deep pages with the most links - probably external links - pointing to them.....

Of course - that's just my interpretation.....

[Robert_Charlton]

Originally Posted by Chris_D

I interpreted Matt's comments - "its all algorithmic" and "expose useful links" as indicating the links selected to be displayed are most likely going to be the deep pages with the most links - probably external links - pointing to them.....

I'd always assumed the links displayed had to do with which other pages would rank for a given query, since the effect itself is query dependent, but on deeper analysis I'm seeing that's not the case.

While being dominant for a given query is what's triggering the effect, the pages returned are not in the same order as the "More results from..." pages, which do in fact seem to be the other site pages in order of how they'd rank for the query.

I have two sites that show this effect on searches beyond the name of the site. On both, the deeper links displayed are the same whether the search is for [keyword1 keyword2] or for the company name.

On one of the sites, the deeper links correspond with what I have good reason to think the inbound linking situation is. On the other site, I'm not so sure.

Also, as I mention in the post I link to above, Google previously had used image link alt attribute, but they're not exactly doing that now.... Where they had used the alt attribute of an image link ("Great Widgets") that links to widgets.html to generate the link text in the serps, they're now displaying the link text as just "Widgets." That one-word link anchor doesn't exist anywhere on the site, nor, to my knowledge, from any outside source. Also, there's another graphic link to widgets.html at the top of the site with an empty alt attribute (" "), and they've ignored that one.

I think this does suggest that Google is looking at more than link on a page... and that it's possible that there's some sort of cross referencing and/or overlap among onsite anchor text, inbound links, and page file-names happening. I could of course be reading too much into this one example, but thought it worth exploring.

[vayapues]

I just got this distinction for the term "astronomy for kids"

I have had the # 1 spot for years, but just in the last 20 days did I get the sub-topic listings.

Not only that, but I also have the next listing under it. Kind of cool. This website has been on the Internet for nine years.

Interesting to read up on why and how it is all determined. Great discussion topic, thanks for bringing it up.

[rustybrick]

More for you at http://googlewebmastercentral.blogsp...sitelinks.html and http://www.google.com/support/webmas...334&topic=8523

[mphung]

All I can confidently add based on my own sites is a) that this does not only happen with clearly identifiable brand names (because there are examples of searches that may or may not be names), and b) it is not based on directory structure, although it may have something to do with internal link structure.

(In a logically built site, the directory structure and the internal linking structure would be the same or similar; but in the example I have in mind it's clearly not the directory structure that determines which links are getting displayed)

AFAIK, it only happens to the listing in the #1 spot. Anything below that has no chance of seeing the additional links displayed.

[dannysullivan]

I've also done a blog post on it: Google Sitelinks: New Name For Those Links Under The Top Listings, which has a little info on the background on when they began and a screenshot.

[Webmaster T]

Matt Cutt's explaination contained two things I believe are important to figuring out how they decide what links to include.
1) it's algorithmic not human edited
2) useful links

So... IMO, it is based on analysis of users sessions. In other words when someone queries with a general term like Ford, what link was clicked in the results and what was the outcome of the click ie: was it what they wanted?

By aggregating the sessions they know what types of info users were looking for and which pages provided the most satisfaction for the user of a general brand query. The Webranking query was the most interesting and did seem to cover the topics which most users would be looking for from a general "brand" query. The GM query is less indicative of it being user session based. IMO, it definitely has little to do with the site itself, it's about the user's wants ie: what do most users find useful from a general "brand" query. That IMO, can only be determined by session anlysis.

[Robert_Charlton]

Originally Posted by Webmaster T

So... IMO, it is based on analysis of users sessions....

On one of my examples, at least, I'd be surprised if this is true.

Btw, the link to Danny's SEW blog post above is broken... Here it is...

Google Sitelinks: New Name For Those Links Under The Top Listings

Incidentally, some of the posts on this thread... in fact all prior to Barry's post... were made before we had the name "Sitelinks." Good to know what to call them.

[ryan3prime]

I have been looking into this both for my own information as well as on behalf of Clients. I have several that have this "honorable" mention status in Google and one who wants it.

First of all it is definitely associated with queries that G feels very sure that their determined #1 result is most likely what the person had in mind, so you're only going to see this for company names, maybe even brands.

Second, while I don't doubt that they track and log user sessions to try to improve their results, I do believe that these additions sublinks are more likely selected based upon the most prominent links available from the homepage. They would almost certainly account for how you link to these homepage "runner-ups" from the rest of the webpages on the domain.

So, if I want this for a client that didn't have it already, I would wrap my strategy around ensuring that the pages I want as sublinks are as closely associated with the company name, domain name, brand name, etc as the homepage or top result that I already have at #1.

It's a pleasure to join your community finally and I look forward to discussing practical matters and confounding issues with each of you!

-Ryan Turner

[vayapues]

Originally Posted by ryan3prime

First of all it is definitely associated with queries that G feels very sure that their determined #1 result is most likely what the person had in mind, so you're only going to see this for company names, maybe even brands.

Not entirely true. The term 'Astronomy for Kids' is neither the name of my company, nor a brand name. In fact it is a very general term.

[ryan3prime]

Actually, I had checked out your example, and you're right, it is an exeption to my statement.

I believe this is due to your stated time online of over 9 years, a truly reliable resource as the WWW is concerned. I have noticed Google's preference for mature websites for a long time now. Of course this carries other benefits like being #1 for a search means people are more likely to pick your site to link to as a resource for the topic, more likely to be included in long-standing directories for the topic, likelihood that you have enhanced the domain with additional related materials to the topic, etc.

So it isn't necessarily a brand or company name but the clarity of the quality of the result itself, signified by a history of #1 positioning, still stands as a necessary platform for earning this distinction.

-Ryan Turner

[Webmaster T]

query=astronomy+for+kids

for, is a stop word so is basically ignored, what was real interesting was quoting that brought back the exact same result for that listing in the same position. In this case for whatever reason kidsastronomy.com is the brand being associated to the query, which indicates the domain is also possibly in play for determing what constitutes a brand. It's obviously more then it being a household name.

It could be that the analysis shows that on the term atronomy for kids, the users found this site disproportinately more "useful" then the others in the SERP. No one knows for sure how "what constitutes" a brand is determined. Google engineers know but beyond that it could be any number of things that "constitutes" brand. After all, all domains are a "brand" of sorts, or, should be treated as such by it's owner. It may be that it only seems that brand is a key when it is just something that fits the same criteria, making it seem like it is associated to a brand.

IMO, if they are using on page or link analysis in this algo, it won't be very "useful" for long. That has a lot to do with why I believe it is analysis of something else either in the click stream or session. If it is about users preferences for a brand query, what better place to look then the click stream or search session?

IMO, to use "on page" factors opens the pandoras box and is just giving webmasters another way to manipulate result snippets. IMO, SE's don't do that unless they are pretty sure it can't be influenced by SEOs. Matt's dialogue seemed to indicate links that being "useful" to a user was the motivation for the change. IMO, this is another one of the small things G does to enhance the user experience and differentiate it's "brand" from the others which can't seem to do much more than copy what Google did 5 years ago. ;-)

One way to eliminate "user analysis" as a possible basis for this algo would be to look in the referrers log and look at what pages the user frequented while on the site after querying google with that term. IMO, you'll find the pages that are listed in the SERP are the pages visited most by these users and and more importantly resulted in a posive action being taken ie: read it (duration of visit) or filled in a form etc. What would be real interesting is the order they are in when sorted descending by pageviews. If it is exact to the snippet then.... it's an even stronger indication of it being "user analysis" based.

Some of the discussion during BigDaddy's rollout was about the possible role that clickstream data and conversion data may have in influencing ranking. IMO, if they are using CT data and clickstream data then... this would be a natural upgrade to occur from the use of that data. I would also note that one way people determine the quality of results is when they query a "brand", it isn't buried below affiliates and re-sellers!

I do some work for an ecommerce site that sells a lot of top brands and one thing I've noticed lately is you are much more likely to find the brand at the top then say a year ago. This site is in a very competitive industry and often the brand was not #1 even for a query for the brand.

Not saying it is "user" based, it just makes more sense to me based on Matt's dialogue. He doesn't tell us it is algorithmic if that is a clue to figuring it out. It's obvious that they see it as enhancing the SERPs they are used in why help SEOs degrade it with manipulation?

[pleeker]

I get the impression that, until there's some serious study of Sitelinks, every idea and theory we propose here will have an exception.

Do a search for the rock band U2, and the band's official site doesn't get the Sitelinks treatment. Nor does my site, which is #2 in the SERPs (@U2).

But, do a search for atu2, and my site gets the Sitelinks treatment. I've never noticed this before, mainly because it's the first time I've ever searched for [atu2] spelled out like this.

It makes sense for Google to show News, Pix, and Lyrics. Those are three of the busiest pages and most highly linked (both internally and from other sites) pages I have.

But there's no way to explain how the "Multimedia" page gets a Sitelink. It's one of the least-visited pages on the site, and doesn't have a lot of inbound links. It has no business being shown as a Sitelink - at least none that I can figure out.