Archive for April, 2008

Does Google allow Cloaking with h1 tags?

Friday, April 25th, 2008

Cloaking defined by Wikipedia

Cloaking is a black hat search engine optimization (SEO) technique in which the content presented to the search engine spider is different from that presented to the users’ browser.

From Google’s FAQ section for webmasters you can also find the following statement:

To preserve the accuracy and quality of our search results, Google may permanently ban from our index any sites or authors who engage in cloaking to distort their search rankings.

Webmasters have many different opinions when cloaking can be used and when it should not (you can read more cloaking threads at webmasters.com). In one particular instance I was looking at a website from a Fortune 100 company that I really liked (bcg.com), and it surprised me to find a cloaking technique in their H1 header text.

Are they really cloaking? Yes they are.  A person visiting the website only sees a logo on the top left corner, while the H1 tag is what the search engines index.

How are they doing it?: The company’s cloaking solution uses a simple css technique: they wrap a link around a div (named “logo”) which is then wrapped around an H1 tag.  Here is the html code:

<a href=”http://www.bcg.com” style=”cursor:hand”>
<div id=”logo”>
<h1><span>BCG - The Boston Consulting Group</span></h1>
</div>
</a>

Then they use CSS in order to hide the text behind the logo (so that only the picture logo displays to the user:

#logo {
width: 260px;
height: 111px;
margin: 0;
padding: 0;
background: url(’/img/logo.gif’) top left no-repeat;
text-align: left;
float: left;
}

h1 span {
display: none;
}

Why not separately display the image logo and H1 tags?  Having a visible H1 tag on every page may not look aesthetically pleasant to the end user and some companies would prefer to show just a well recognized logo to their uses.  Cloaking H1 tags under an image logo provides a way to show the logo to the user while still have each page’s H1 tag include their brand name.

Are they alone? No.  Many other Fortune 100 companies (like Qualcomm, Quicken Loans, Ohio Health, Gore Tex, SAS) are also using this particular technique. I don’t believe Google will ban any website from cloaking their H1 tags with their image logo as long as it is clear that they just want to provide a better user experience and not distort search rankings.  In most cases all the companies that I’ve searched and found are just cloaking H1 tags with images for their brand name.

Recomendations on Cloaking Images: Because the sites in question are not already banned from Google’s search index, it’s probably the case that cloaking H1 tags with your brand image logo should be ok and won’t get your site banned.  In this case, it should be clear to Google engineers that you are not trying to change the message you give to people vs search engines,  but want to provide a better user experience.  They keyword here is “should”, so keep in mind that any form of cloaking in the long run is probably not going to be worth the effort - only the search engines know what is acceptable when it comes to what is considered “cloaking” and what is not - so it can become difficult to guess.

Also keep in mind that it’s usually not too difficult for a site to rank well for their brand name because normally there isn’t going to be a large number of competitors with the same company brand name out there.  Keeping a brand text in visible H1 tags, in the site’s title tags and or in backlink anchor text might be a better option instead of cloaking on any level.

These icons link to social bookmarking sites where readers can share and discover new web pages.
  • bodytext
  • del.icio.us
  • NewsVine
  • Reddit
  • Slashdot
  • StumbleUpon
  • Technorati
  • Netscape

Google’s Page Rank - A Population Control Mechanism

Thursday, April 10th, 2008

I get a lot of questions on Google’s page rank meter. I was asked by a friend of mine:

Question 1: Based on the Google Page Rank, the highest rank I’ve come across so far has been with latimes.com (8 out of 10) and nytimes.com (9 out of 10). Besides Google, Yahoo and the rest that calculate page rankings, do you know of any sites that have a perfect 10 ranking?

Question 2: Last week I saw the LATimes.com site ranked 9 out of 10, but this week it’s back to an 8. Any thoughts on why they dropped?

To be short and sweet, I don’t know of any sites that have perfect 10 page ranking. But should you even care?

Page Rank is one small portion of Google’s total algorithm and only looking at one small portion of a larger picture often times doesn’t give you the best point of view. A site can rank in different positions for many different keyword phrases, so PR isn’t going to, in itself, indicate where in the results you will show up for a given keyword - since your positioning is based how a person searches for your site along with many other factors that happen to include Page Rank in the equation.

Well, what does the Page Rank (PR) meter indicate anyway?

You could say PR tells you how “networked” your site is. PR is thought to be a logarithmic function of the number, and possibly the total quality, of in-bound links a site has accumulated. So, the more in-bound links you have, the higher your PR. But it gets harder and harder to climb up the PR ladder since the equation has a logarithmic base.

Page Rank meter, Not Always Accurate

You may have noticed: the Page Rank meter that you might have installed on your FireFox or IE web browser is not always accurately reflecting Google’s internal Page Rank numbers for 2 reasons:

  1. Google updates their PR values once every 3 or 4 months - so the PR value you see for any given web page is not always going to be up to date.
  2. The Page Rank meter has to guess a page’s value some times, for example, the meter PR shows a value on the Gmail interface page (which shouldn’t have a rank since your private email messages don’t appear in public web results) - so we don’t even know which pages the meter is guessing on!

Google’s Page Rank MeterSalt Shaker

With all this said, Page Rank should probably be taken with a grain of salt in that it can be very misleading piece of information.

OK, but how can a site’s Page Rank drop?

A site’s Google PR can drop for a few reasons that I am aware of:

  1. a reduction of in-bound links to the site
  2. a Google engineer manually penalized the site in question for spammy web practices
  3. the meter might actually be “guessing” a page’s PR value to a certain extent, almost like a random number

Is there any value to the PR number of a web site?

Yes. Since the Page Rank number of a web page is updated every 3 months or so, we know we are looking at historical data (an image from the past). For this reason we can usually judge the past activities of a site in terms of in-bound links and how much “link” juice a site receives in general - and if all else equal (the site has not participated in any type of known web spam activity since the last update) then we might even be able to determine the current health of the site - weather or not we want to acquire a link from the site or if we want to link out to the site.

Why does Google make the Page Rank numbers public?

  • When you think about it, Google wants to do everything in their power to stop people from gaming their search engine. So why make an internal number from their algorithm public?? (It’s possible that they don’t even use PR in certain ranking calculations, but that’s another story)
  • Google uses the PR meter so it’s user population will learn to value sites the a high page rank and devalue sites with a low page rank.
  • If they can get you to think in their terms, they can control your fear: what do you think Google does to sites with a healthy page rank that commit “spammy” practices (such as selling links), or anything else they deem as “bad”… they drop the site’s page rank down or even to zero as a scare tactic - sort of like a slap on the wrist and “let that be a lesson to the rest of you!” - even though the site’s rankings, and traffic don’t change - the things they can’t obfuscate with a simple reduction in PR.
  • Google uses the “Page Rank” (PR) metric as a means of controlling the behavior of the population when it comes to judging the “quality” or the ability of a site to rank in the search results - that’s why they make the information public.
  • The PR meter helps prevent people from spamming their search engine, so they can generate cleaner web search results, make it harder to perform organic SEO and drive more money to their adwords program.

And no, other than Google’s home page, I don’t know of a web page that has a perfect 10 PR, do you?

These icons link to social bookmarking sites where readers can share and discover new web pages.
  • bodytext
  • del.icio.us
  • NewsVine
  • Reddit
  • Slashdot
  • StumbleUpon
  • Technorati
  • Netscape

“Home” vs “Homes” - which keyword drives better traffic?

Wednesday, April 2nd, 2008

Homes

I recently received a question from a client asking which sub domain would be the most beneficial to create for their new real estate web site, either going with “home.site.com” or “homes.site.com”:

I have one specific question. It was suggested by a member or our marketing staff that for Real Estate sites, “home” gets better results than “homes”, particularly as part of a url. For example “home.site.com <http://home.site.com>” is better than “homes.site.com <http://homes.site.com>“. …. ‘and it wasn’t even close’, I was told.

Can you guess which keyword brings in more interested traffic? Let’s find out:

Search Volume of “homes” versus “home”

According to Google’s trends tool, “home” receives more searches than “homes” by about 2 to 2.5 times as much. At first, the larger search volume would lead you to believe more (in blue) is better…

home vs homes

Wordtracker also reports that “home” has a larger search volume than “homes”.

Google’s adwords keyword tool also reports the same trend:
https://adwords.google.com/select/KeywordToolExternal (just enter the two words)

Keyword discovery also reports similar data (768196 for “home” vs 135921 for “homes”)
http://www.keyworddiscovery.com/search.html

So there really are many more searches for the singular than the plural version of the word - but there’s more to the story here…

Correlated searches of “homes” vs “home”

As we’ve seen, there are more searches being reported for “home” than “homes” but after looking at keyworddiscovery’s data, for example, “homes” seems to be more correlated with related searches such as “homes for sale” or “mobile homes“, etc., while the keyword “home” includes a lot of search queries looking for things like “work from home” or “home garden“, etc.

Homes search results on keyworddiscovery Home keyworddiscovery search results

It would appear that if you were going to start a real estate related web site, “homes” might be a better choice in terms of traffic quality. Even Google seems to thinks so, they have a link to “Find results for homes in Housing” right above their main search results.

These icons link to social bookmarking sites where readers can share and discover new web pages.
  • bodytext
  • del.icio.us
  • NewsVine
  • Reddit
  • Slashdot
  • StumbleUpon
  • Technorati
  • Netscape

Entries (RSS) and Comments (RSS).