Rank improvement by changing URLs

Back in December 2011, almost one year ago, friendly URLs were introduced in order to improve CTR. The old parameterized URLs were indexed, but even when ranking highly, the CTR was far from impressive.

The switch to URLs with clear and appropriate keywords to provide a logical structure has helped encourage additional visitors to click, and there has been a corresponding increase in rank. Of course, whether this is a causal relationship is up for debate – but either way, the statistics are interesting…

The following graph shows the steadily improving position in SERPs for the friendly URL since their introduction at the start of the year:

Rank improvement from friendly URLsThe chart shows a starting rank of around position 50 in Google back when the URLs were changed. As of last week, the URL is in the top 20 and has featured in 5th position regularly in recent weeks.

This positional change is not due to link building to the page:

No new backlinksThe backlink chart above suggests that no new links are being built in easily discoverable parts of the internet. The total number of backlinks reported has remained zero throughout the period of the rank improvement.

Similar improvement in rank is seen across other friendly URLs.

One explanation is that Google tests URLs for CTR. Those exhibiting favourable metrics e.g. low bounce rate, time on site before returning to search again could be used to indicate results that are the most valuable to search. Those results should be promoted to better positions. This promotion in search results takes time.

Even before Penguin, there have been alternative approaches (beyond simply building keyword rich anchor links) to improve rank. Improving the appearance of snippets to increase CTR is one of those techniques.

 

Is link building dead ?

The appearance of fresh links may provide a useful signal to Google – to test a URL higher up in the SERPs – but if that URL doesn’t engage the visitor, then other metrics will be poor and the URL is more likely to slip in the ranks than climb.

Heavy linkbuilding may simply hasten the decision to demote a poor performing site, based on the engagement metrics of the URL targeted.

What do you think?

Please leave a comment below.

 

 

Enhanced by Zemanta

Stem terms not Head terms: Latent semantic indexing

Keyword research has been the foundation of search engine optimisation plans for over a decade. The identification of high frequency search terms has driven many clients to expect results in terms of rank rather than ROI.

How many times has a client called to ask what can be done to get term X into first position in Google ?

How many times do you ask the client why they want that particular term in 1st place ?

The most often reason given (following a stunned silence) is that it is “clearly an important term for their business, it drives a lot of traffic”. Now, traffic may be important if you are running an affiliate site, or a business running advertising, where your revenue is governed by clicks. However, driving a lot of traffic does not equate to driving a lot of sales on eCom sites, or service industries or sites where the content simply does not match term X.

These terms are often referred to as ‘vanity terms’ or ‘glamour terms’. The keywords that the business owners ‘believe’ are important to their business when the truth in the data shows it is often the longer tail terms, brand traffic or other channels in combination that actually drive the sale.

The industry has fueled mis-understanding by calling short, highly searched keywords ‘head terms’. The phrase ‘head term’ carries an implication that the keyword phrase itself should be found at the top of search results.

…so we have a situation here…

Businesses are calling for ‘vanity terms’, the search industry is confusing matters further by referring to short highly searched phrases as ‘head terms’ as if it was valid that either should be ranked highly…

For example:

A small estate agency on the Isle of Man could get thousands of visitors per day if it were at the top of Google for the term ‘houses for sale’.

However, it would be much better for the estate agent to be at the top of Google for the phrase ‘houses for sale in the Isle of Man’ or even ‘houses Isle of Man’. Neither of those terms have huge search volumes, but both are imperative for an estate agent based on the Isle of Man selling houses! Visitors arriving on their site are pre-qualified and much more likely to make an inquiry.

For illustration, here’s a list of the terms of real importance to our fictitious agent

  • houses for sale in Douglas
  • houses for sale in Castletown
  • houses for sale in Peel
  • houses for sale in Ramsey

and the long tail might be:

  • 2 bed houses for sale in Douglas with garden
  • 4 bedroom houses for sale in Ramsey with garage

Rather than confuse clients by calling the keywords ‘houses for sale’ the head term, a much more accurate terminology is concept of word stems.

Key phrase Stems not Head Terms

In linguistics, a stem is the part of the word that never changes when considering the various forms of the word. For our purposes, a key phrase can been seen as a compound word, or extended stem about which we build the long tail tail terms.

In the example above, the phrase ‘houses for sale‘ becomes the stem phrase.

We can then build variants around this stem, to get the longer tail terms that are important to the business.

By talking about ‘houses for sale’ as a key phrase stem, there is no longer an implicit expectation for the ‘stem’ to reach number one in Google. It is much clearer that this is the building block around which we can optimise the site.

If we revert to talking about ‘house for sale’ as a head term, then you immediately re-introduce the idea that it is in some way important for this phrase to be at the top.

Head Term implies top and important

Stem Term implies simply a building block

As SEOs, we must explain the importance of the long tail and the concept of LSI to our clients.

At the very beginning, we should refer to words essential to the long tail search, as ‘key phrase stems’ or ‘stem terms’ instead of confusing a client using the phrase ‘head terms’.

Stems and Semantic Indexing

Exact matching terms (either on page or off page) are becoming less important over time and indeed, if over-used, exact matching terms can be detrimental to a site’ performance. Exact matching anchor is widely acknowledged as the target of the Google Penguin updates – repeated anchor text is easily identified by an algorithm.

As algorithms improve, more sophisticated connections are analysed to determine the latent semantic content of sites. Content, context and connections all play their part in deciding the relevance of a site to a search query.

By thinking in terms of stems, you change the way you think about connections too. A stem is a building block, from which ideas can grow and flow. Ideally, the stem is the distilled essence of the business, the idea behind the search term rather than the search term itself. There is often no need for the stem phrase to rank highly for a site to be successful.

In our example, the term ‘houses for sale’ may have enormous search volume. For a real estate business in the Isle of Man, enormous search volume might sound highly attractive.

Ask yourself: Who makes the agent think this way?

Are you surprised that after they read your report calling ‘houses for sale’ a ‘head term’ with 3,000 searches per day, that they believe they should rank 1st for this phrase now?

If you’d explained that the ‘stem phrase’ for their business was ‘houses for sale’ and the important converting long tail terms were built around this stem, then your client is more likely to understand why ranking for the term ‘houses for sale’ in isolation is simply pointless…

They’d also see that ‘houses for sale in the Isle of Man’ was actually the key phrase for their business.

To bake your noodle, the longer tail terms are actually the ‘head terms’ for most businesses!

The short tail phrases are the ‘stems’ around which longer tail phrases, semantic connections and ultimately ROI and a successful business is built.

PPC is different

Most PPC managers will know that AdGroups containing exact match keywords with high conversion rates are a good way to limit bidding costs and keep an account under control. The trouble is, converting longer tail variants may have never been seen before and are therefore unknown – adding exact match long tail keywords to a campaign is near impossible to achieve.
A good summary of the issues facing PPC when selecting exact match terms are discussed here

Enhanced by Zemanta

Schema.org examples showing search result snippets in Google

Schema has been big news for SEO for quite a while now, yet very few people are using it. Whatever the project I’m involved in, I make sure schema forms part of the basic on-site recommendations. Why wouldn’t you take advantage of improved snippets, especially when all that is required are simple html decorations.

 

This post shows a few examples of the impact of using schema for improvements in search result snippets in Google.

Example MusicRecording Schema

Music Tracks in search results
Music Tracks in search results

The MusicRecording schema has been used to enable the rich snippet above. The code for the MusicRecording schema was as follows:

MusicRecording Schema
MusicRecording Schema

 

Example Recipe Schema

hrecipe is currently the most popular mark-up method, however the schema.org recipe can help dominate the SERPs.

This is how the common hrecipe microformat appears in SERPs:

Using hrecipe and appearance in SERPs
Using hrecipe and appearance in SERPs

In contrast, using the recipe schema can help occupy more space, the more ingredients you add…

SchemaRecipe helps dominate SERPs
SchemaRecipe helps dominate SERPs

The code used to achieve the result above is as follows:

Code for SchemaRecipe with several ingredients
Code for SchemaRecipe with several ingredients

It’s easy to see how schema could be used for tactical snippets.

Example Movie Schema

Movie Schema in SERPs
Movie Schema in SERPs

The code to create a snippet for movies as above is as follows:

Code example for the Movie Schema
Code example for the Movie Schema

The movie SERPs show the beginnings of how decorated data can be used to create readable text. ‘Directed by xxx, starring xxxx’ is the first step towards producing content created entirely automatically from an enriched data source.

Example of TV Series Schema

The TV Series schema offers ratings information:

TV Series in SERPs
TV Series in SERPs

To create the TV snippet above the following code was used:

TV Series Schema code
TV Series Schema code

If they’d included the optional TVSeries schema information above, then the snippet could have been better still.

 Example Event Schema

Events and news get special treatment in Google (fresh content), so if you want to get ahead in SERPs, then add a properly marked-up events page to your site.

Using vevent, the search results will look like this:

Using vevent in serps
Using vevent in serps

The code for vevent to create the snippet above:

The vevent html code
The vevent html code

Very few sites currently use the schema Event markup, here’s an example from the SES site:

Schema Event in SERPs
Schema Event in SERPs

and the code used to create the snippet above is the result of a minimal implementation. Schema.org offers many types of richer event markup, so this entry could have included much more information.

Code to create Event Schema
Code to create Event Schema

 

Sites are still not using schema.org to its full potential, despite the major search engines announcing that they support this format as the standard. There is substantial scope for all sites to incorporate pages with schema in order to improve appearance in SERPs and encourage the click.

Why is this taking so long to catch on?

 

 

Enhanced by Zemanta