Typical proper use of the rel="canonical" tag (via SEOMoz.org)

One of the fun problems we have working at WhatClinic.com is trying to organise the millions of pages that result from listing tens of thousands of clinics in thousands of locations for thousands of treatments.

Our search results pages list up to 12 clinics at a time, and when they’re full they offer a great user experience. Lots of choice and lots of information is presented along with a simple way to contact whichever of the clinics takes your fancy.

However, not every combination of clinic type + location + treatment will have a full page of results. In fact with only a little knowledge you could probably guess the URL of a page with no results on it. The obvious solution to these empty pages is to return a 404 response code and not to link to the pages internally, minimising the chance that they’ll be found by users or search engines alike.

What’s Right For The User?

Add one clinic to the page though and we’re left with a quandary. Is this really a useful page for a user? Wouldn’t they like more choice? We know for instance that pages with more clinics on them have a better conversion rate, so would we be better off sending users to a “parent” location page instead, i.e. a location that contains the smaller location but should have more than one clinic on offer?

Another option available to us would be to fill the rest of the page with 11 of the nearest clinics to the location (which could be tens if not hundreds of miles away in some cases), but this would massively increase the duplication of data served across the pages on our site as clinics’ listings would appear in far more locations than they currently do.

Similar Pages – The Rel=”Canonical” Solution

We decided we’d like to see what effect the first option had, i.e. sending the users to a parent page, but we were uncomfortable with 301 redirecting every page that only had one clinic on it, so we decided to try a slightly softer approach.

Having read an article on SEOMoz about using the Rel=”Canonical” tag to get more than one keyword to rank for a given piece of content, we decided to try what we thought was quite a clever scheme that would serve the user and the search engines.

We would put a Rel=”Canonical” tag on our search results pages with only one clinic listed, and we’d hope to send people searching Google for Place A to the search results of Place B, which would contain the search results for Place A and more, giving the user a better choice.

Anchor Text Isn’t A Very Strong Ranking Signal For Pages With A Rel=”Canonical”

Unfortunately for us, the experiment hasn’t exactly gone to plan. We were cautious and only put the Rel=”Canonical” links on a subset of our one result pages, but even still we have enough data to see that for now at least none of the Place B pages are ranking for Place A keywords.

Of a sample set of 20 one result pages with a Rel=”Canonical” tag, 14 have been crawled and no longer appear in Google’s index, and searching using the “Place A” keyword for these pages doesn’t return the Place B search results page.

You might think, well Google have decided that the Place A and Place B pages aren’t sufficiently similar to be a valid use of the Rel=”Canonical” tag, and you might be right, but the fact that original Place A URLs are no longer appearing in the index seems to counter this supposition.

More likely it seems is that the anchor text of the links pointing at Place A pages isn’t a strong enough signal for the Place B pages to rank for keywords based on “Place A”.

Back To The Drawing Board

So it looks like we’re back to square one on this particular problem. I think the next thing to try is the option discussed above where we fill out the search results. It seems like a good thing to do for the user, but I am slightly worried about diluting our content by potentially overusing it. We’ll be sure to keep you posted about the results when we try it out.

Have you run any experiments with the Rel=”Canonical” tag yet? For what purpose, and what results did you see? Let us know in the comments.

We’ve just added a new set of navigation tabs to our search results pages after A/B testing them for a week. They didn’t improve conversion by any significant amount but by adding them we give ourselves the flexibility to add new ways of slicing our data, and hopefully capturing more traffic.

First of all, we’ve moved the map link which was on the top right hand side of the page into one of the tabs. This has already increased usage of the maps pages on the site. Secondly, we added two new types of pages, one which pulls together all the reviews for the clinics in the search results, the other all the prices.

These two tabs are brand new pages with their own URLs and are SEO’d to capture traffic relating to reviews and ratings of clinics and treatment prices. This is an SEO experiment in itself. All of the data on these pages exists on other pages. For instance, all the reviews of each clinic exist in their own profile, but these new pages pull together what we think are a useful set of reviews so they can be compared in one place. The same goes for the prices.

Here’s what a normal search results page looks like: Dentists in Dublin

Here’s the reviews page: Reviews of Dentists in Dublin

And here’s the prices page: Prices of Dentists in Dublin

Time will tell whether or not the search engines think these pages are useful, but we’ll be keeping a close eye on how their traffic and usage grows in the coming months.

Have you any experience with re-slicing your own data? How did it work out for you? Leave us a comment below.

Google Search Result For Cosmetic Surgery Ireland

Google have started to add extra information to their search results in order to give users a better idea about what section of a website they’re about to click through to. This makes more sense in some situations than others. For example, in the search result pictured above the new site hierarchy link “Plastic Surgery Clinics” which appears after the “www.revahealth.com >” in green doesn’t really add any information, seeing as the same text is included in the main link to the page.

However, imagine if the page title (and Google’s main link text) was “Dr. Michael Smith, 4 Main Street, Crosstown”. In this case the site hierarchy link would probably be something like “Dentists”, which would give you a good context about the type of page you were going to click through to, telling you Dr Michael Smith is a dentist rather than a GP or a surgeon for instance.

Google Search Result For Dentists Ireland

These site hierarchy links have only started to appear in the last week, and so far only for some of our search results. It will be interesting to see how common these become and how they’ll evolve. Have you seen these for your site yet, and are they any help?

[Additional information direct from Google here: Site Hierarchy Links]

The main goal of all search engines is to return results from the sites that are the most useful and relevant to a user query. Sites with country-coded top-level domains, such as .ie, are associated with a geographic region, in this case Ireland. By default, and all other things being equal, sites with a .ie domain are interpreted as more relevant for users in Ireland. For generic top-level domains such as .com Google allows site owners to manually assign a preferred country in their Webmaster Tools.

As our .com domain’s content covers many countries we can’t use this particular feature to optimize the site for a particular region. Instead, we were hoping that by redirecting our Irish and UK pages to their local domains that those pages would benefit from a boost in search results positioning on Google.ie and Google.co.uk as being targeted at their local audiences.

The initial results were terrible. The redirected pages dropped in the SERPs by about 3 positions. As our Irish pages previously ranked better than our UK ones they suffered even more. (Ireland was our first local area of interest and those pages were in the index much earlier.)

After a few weeks our results moved slightly in the right direction but were still behind the control group. We decided to let the experiment run for longer, almost four months in total, to be sure that the conclusion drawn was correct. In the end we turned off the redirections, just last week in fact. The experiment was successful in terms of achieving measurable and unambiguous results, but unfortunately they were very disappointing.

Our redirected .com pages disappeared from the index and over time their corresponding .ie and .co.uk pages were indexed instead. However, they never reached the search results positions we had had before with our .com pages. We speculated that our .com domain had more trust and authority than our country specific domains and that’s why pages from RevaHealth.com ranked better even in Google’s country specific search engines.

Now, just one week after undoing the redirects, most of the .com pages have regained their positions from before the experiment, which we think goes some way to validating our speculation. Our generic .com domain, which is crawled much more frequently than our local domains are, does seem to be a more trusted site, not only by Google.com but also by Google.ie and Google.co.uk.

Have you run any similar experiments that you can share with us?

The whole reason why search engines work as well as they do is because of the huge amount of effort that publishers put into SEO daily. That’s right – without SEO the search engines wouldn’t work.

Why?

Well, what a search engine tries to do is understand the meaning of a page and subsequently display it in the search results when someone types in a query that indicates that they are looking for that content. This is phenomenally difficult to do; even with the best brains on the planet it is currently impossible.

The reason for this is that machine learning isn’t nearly advanced enough to be able to understand at the levels that humans are able to understand.

So the search engines rely on publishers to shape their content so that they can better understand its meaning. Back when the search engines where pretty unsophisticated this literally involved telling the search engine using the meta keywords and meta description tags.

As the search engines and spammers have become more advanced the means of conveying meaning to the search engines has become less overt, but no less important.

RevaHealth.com definition of Search Engine Optimization:

SEO is the science of narrowing the gap between the search engines’ understanding of a page and a human’s understanding.

You will notice from the above definition that this doesn’t include link building. That’s because we regard link building as PR. Just as old school PR was about getting your company correctly positioned in newspapers and on radio, new school PR is the same thing online.

So what can happen if you build a website that solely targets users and not search engines?

The following examples are completely legitimate ways of constructing your website that will have no negative effects on your end users, but will completely ruin your chances of being distributed through the search engines.

Let’s say you are building a listing site similar to RevaHealth.com

1. You use parameters instead of hard URLs.  If you construct your parameters without thinking about SEO  you can easily create a website that the search engines simply won’t index.
2. You use JavaScript to dynamically call in the content similar to Kayak. This can provide a very useful end user feature. However, if coded without reference to SEO it will result in the content not being indexed.
3. You use a template for each page that results in the search engines thinking that each page is a duplicate of another page.
4. You include useful additional information for the user (for example from Wikipedia) that is not original.   At best this will result in the search engine marking down your pages and at worst they might list you as a spam site and exclude your entire site from the index.
5. You have a infinite number of combinations of search results available resulting the search engine giving up before they have crawled all of the relevant content

In Summary

Search engines are the number one means of distributing your content. If you develop a website without considering the SEO implications of your decisions, you are effectively giving the search engines two fingers. Don’t be surprised if they give you two fingers straight back.

We often get asked about our experiences with SEO, usually about what worked for us and what didn’t. Keywords are a topic that come up again and again. The video above comes from Matt Cutts, the head of Google’s Webspam team, and it neatly summarises the fact that when it comes to Google’s main search product, meta keywords are completely ignored.

That should be the end of that then. But it isn’t.

A lot of people seem to get mixed up between keywords and meta keywords, thinking they are exactly the same thing, so the video above might be taken by some people to mean that ALL keywords are ignored by Google. That just isn’t the case.

The word “keywords” can mean many things interchangeably, but for the purposes of this blog post, let’s  refer to a set of keywords as the most important words, phrases or acronyms that your potential visitors will use to find a particular page of content on your site.

For these potential visitors to find your page, the keywords they use, or their synonyms or related keywords will have to appear on or pointing to the page somewhere. These keywords can  appear in:

• the body text of your page
• the URL of the page, including the domain name
• the page title of the page
• the meta description of the page
• the H1 tag on the page
• the alt text of images on the page
• the links to the page

And that’s just off the top of my head. They can also still appear in your meta keywords, even if Google currently chooses to ignore them THERE.

You should still be doing your keyword research for each page (or set of pages) and using those keywords you identify as being the most important in the page elements listed in the paragraph above. Google even offer some good free tools to help you find out what keywords to use. I’d recommend Google’s Search Insights, Google Analytics, their free AdWords Keyword Tool, and even Google Trends. Add your own favourite keyword research tools in the comments below.

The SEO problem with duplicate pages is that Google doesn’t want to clog up its index with a whole bunch of duplicate content, so it tries to cull the duplicate content from its index. In our case it only includes cosmetic surgeons in London in its search results.

In the past we included machine generated text on each page, in an effort to describe on the page in a way people could easily understand what the content was without having to write tens of millions of descriptions by hand. However, because this particular block of text was quite similar from page to page, it hindered us rather than helped us in relation to duplicate content. So we set about trying to find a way to increase the originality of each page.

On a side note, it is possible for you to take control over your own duplicate content and to tell the search engines which page is the original or most important version of a page rather than letting them make that decision for you. You can use either canonical URLs or 301 redirects, something we’ll discuss in another blog post. For now, this is something that we do already, but as the pages are actually valid, non-duplicate pages for our visitors, we think that this shouldn’t be necessary.

So, going back to looking at how to increase the originality of the content on our pages, we took our search results for Dentists in Mexico as our test bed. For 50% of the locations in Mexico we added 2-3 paragraphs of location descriptions taken from Wikipedia. Wikipedia has relevant content that can be re-used on other sites thanks to the GNU Free Documentation License. The link to the original source of the text was included underneath.

We were hoping that syndicating content from Wikipedia could alleviate the duplicate content issue along with giving our visitors a better experience. We let the test run for three months.

The Results

Although our results shouldn’t be regarded as complete, we found that the inclusion of Wikipedia content on our search results pages had no effect on whether the page was included in the main Google index.

However, we also found that all pages with Wikipedia content that were already in the search results dropped by around 3 positions, while all control pages gained on average 2 positions!

Search engines want and reward original content. It is known that Google uses document similarity techniques to keep searchers from finding redundant content in search results. Our experiment left no doubt about it. I only wonder how will Google solve the problem in the current large scale web syndication era if it is possible to find 5 exactly the same articles on the top 5 sites in the results for many of Google searches, e.g.

http://www.google.com/#q=Get+Motivated+to+Create+New+AdSense+Content

With just short of 80,000 pages from RevaHealth.com in Google’s active index there are an awful lot of ways that visitors can find pages on our site. Looking at the keywords they searched for before arriving on the site should be an easy way to gain an insight into what our visitors want, but the sheer volume of unique keywords arising from a long tail strategy makes this much harder.

For example, in May 2009 there were 83,153 unique keywords that brought people to RevaHealth.com, and the average number of words (fragments) in each keyword was 3.88. Compare this to the 2008 global average of 2.2 words per query and you start to see how different our traffic is. The graph below helps illustrate the problem – there were 13,097 fragments that were searched for just once in May, while there was one word that was searched for 17,130 times!

Keyword fragment usage rates May 2009

So what was this most searched for fragment? Not surprisingly for a site organised by locations, it was “in”. Also unsurprisingly for RevaHealth.com, numbers 2 and 3 were “dentist” (12,837 times) and “dental” (10,041 times) respectively. The difficulty however comes with how quickly these numbers drop off. For instance, our hundredth most popular keyword fragment was “leicester”, with just 379 searches in May. In fact, you may have noticed that the above graph’s vertical axis uses a logarithmic scale just so you can see any detail!

The point of all this is that using a long tail strategy to attract traffic does make it harder to spot the underlying trends by mining your keywords. Some tips I would suggest are:

1. splitting keywords into fragments and aggregating – this at least helps to point out the most repeated individual words
2. create groups of keyword fragments under different themes or meanings and measure their collective popularity – this can help to identify broader trends i.e. which is more important to your visitors – price or quality?
3. watch out for rising trends by looking for unexpected keyword fragments appearing high on your list – these can help to identify timely sales opportunities

I’d also be interested in hearing how you go about analysing your keywords to identify important SEO terms, visitor trends and sales opportunities, so leave a comment and let us know.

There are a few reasons why we optimise RevaHealth.com for long tail key phrases. Firstly, it’s much easier for us to rank well on search engines for rare key phrases because of the lack of competition. For example, our site isn’t listed within the Top 10 Google SERPs for ‘dentists’, while for ‘dentists dublin’ we are in position one. Being able to rank so well for these long tail keywords guarantees us a large slice of the traffic these keywords generate.

Secondly, in order to target the long tail, we have created a lot of pages, which means we end up with a lot of pages indexed (39,900 so far), which gives us great visibility. What’s more, this increased visibility is very valuable, growing our audience and earning potential revenue. What’s more targeted visitors seeking very specific information tend to useg long tail keywords and as a results are more likely to convert or click on contextual ads, increasing our revenue per visitor.

Discovering which long tail key phrases are relevant to your business isn’t difficult but it is time consuming. There are a number of great free tools available from Google: Google Keyword Tool, Google Insights For Search, Google Trends. Sometimes however the simplest way to get good information is to actually publish pages with unique sets of keywords and see whats gets traffic. We rely on this method of generating content as we continue to add new treatment categories and destinations on a regular basis.

A recent Google study on search engine users’ behaviour provides more evidence that back up the theory of the long tail. According to the research, as time passes, people are using longer and longer phrases in the search box.

The 2008 average is 2.2 words for query, while in 1994 it was only 1.2 words. Some sources mention even higher numbers, for example, the Nielsen Norman Group cite 3.1 words per query. And this number still grows.

These are examples our search terms that have directed traffic to RevaHealth.com:

• how much does braces cost in mexico
• anyone had liposculpture with dr robin van look?
• board-certified plastic surgeon in poland
• list of dentists in singapore dental clinics
• effective face lift in the philippines

The long tail is, then, an undeniable opportunity to attract more visitors. Targeting multi-word search queries, optimising a website and creating content to match the long tail, allows us to win potential users and buyers. The down side is you need to create a lot of pages with unique and valuable content so that the search engines index them.