Wednesday, September 29, 2010

Search Group - Perth SEO company - www.SearchGroup.com.au

Hi all,

Just an update to let you all know that I am now working as a Search Engine Optimization specialist with Perth SEO company Search Group.

Search Group is a leading Perth SEO Company specialising in website search engine optimisation (SEO), Search Engine Marketing (SEM), and online marketing activities with a large and growing portfolio of Australian clients.

Address: Suite 90, City West Business Centre, 102 Railway Street, West Perth, WA 6005

T: 08 9278 8899 
F: 08 9278 8890
E: matt@searchgroup.com.au
W: http://www.searchgroup.com.au/

So if you would like a free introductory SEO training session, website appraisal and review or just simply to discuss your search optimisation requirements to improve the visibility of your website on Google, Yahoo! and Bing then please drop me an email.

Monday, September 27, 2010

404 Error Pages... have you customised yours?

I generally recommend for client websites to include a customided 404 Error Page (page not found error alert) for users who mistype or arrive at a broken link. Best practice indicates that a redirect should be provided back to the website home page or site map. I believe that the custom error page has become quite a fashion statement and feature of discussion. Though you can set a meta refresh to redirect the user after a set period of time (20 seconds) to the home page, I prefer to include the site map or site search feature on this page for usability improvement to the users experience, that clearly indicates that they are presented with a page that indicates that they have come to this point as the page that they were trying doesn't exist, and here are some options to assist in getting to where you were going.

It is important to note the most common reasons that visitors to the website land on a 404 page:

• A mistyped website address or URL (or an out of date bookmark or favourite)
• A search engine link that is out of date
• An internal broken link that the webmaster or web developer is not aware of.

Saturday, September 25, 2010

Factors to consider for local search SEO - using GEO tags

This post combines information on what GEO tags are, how they can be implemented, and factors that may contribute in my opinion to improvements in local search for Search Engine Optimisation.

Firstly the Geo Tag Elements

<META NAME="geo.position" CONTENT="latitude; longitude">
<META NAME="geo.placename" CONTENT="Place Name">
<META NAME="geo.region" CONTENT="Country Subdivision Code">

 e.g.
<META NAME="geo.position" CONTENT="-31.9667;115.8167">
or
<META NAME="geo.position" CONTENT="-31.9667;115.8167">
<META NAME="geo.country" CONTENT="AU">
<META NAME="geo.region" CONTENT="AU-WA">
<META NAME="geo.placename" CONTENT="Perth">


Element Description

Longitude:
Longitude is conventionally given in degrees of arc relative to the Greenwich Meridian, a great circle passing through the poles and Greenwich in London, England. Longitude is either qualified as East or West, or given as a signed numeric value of degrees East. The geo.position tag uses a signed numeric value.

On a Mercator map, Longitude is plotted left to right as the X coordinate, while Latitude is plotted bottom to top as the Y coordinate, with (0,0) in the middle. Thus, positive values for Longitude correspond to locations East of Greenwich, e.g. in Asia. Locations West of Greenwich, e.g. in the United States, correspond to negative values of Longitude.
(Latitude;Longitude) in quadrants

Latitude:
Latitude is conventionally given in degrees of arc relative to the Equator. Latitude is either qualified as North or South, or given as a signed numeric value of degrees North. The geo.position tag uses a signed numeric value.

Region:
The geo.region tag is taken from a controlled list, and may be used for resource discovery. It may also be used as a bounding check to validate the geo.position tag.

For the geo.region tag, the ISO 3166-2 country subdivision codes are used. These codes are formed from the "domain-name" 2-character codes, for example iso3166-countrycodes.txt, together with a regional code. The codes may be found here, while the GeoTag Generator includes a 2-stage script to pick appropriate codes.

For the USA and Canada, the subdivision code is the familiar 2-character state/province abbreviation. See for instance United States Postal Service, Official Abbreviations - States and Possessions and Canada Postal Guide (Province and Territory Symbols)

Country:
If the regional code is not known, the 2-character country code may be used instead.
ISO 3166 codes are reproduced with permission from ISO

Place Name:
The geo.placename tag is provided primarily for resource recognition; it is anticipated that this field be harvested by automated agents and presented to the user in search engine results in a similar manner to the description META tag. This field is free-text, and typically would be used for city, county and state names. It could, however, be used for resource discovery, particularly if names from some controlled vocabulary such as the Getty Thesaurus of Geographic Names is used.

Required Accuracy of Position
The accuracy with which positions need to be determined is largely determined by the character of the resource being described.

While a position given in a gazetteer or thesarus such as the Getty Thesaurus of Geographic Names is convenient, and in many cases adequate, in other cases it is clearly not. For instance, it is probably sufficient to place a branch office of a multinational corporation on a map of the world, or a map of a country. It is, however, clearly insufficient to place a public telephone, filling station or fast-food restaurant. In these cases, position must be measured or researched to greater accuracy.

Accuracy of Elements
Properly expressing the accuracy of geographic positions is too complex for the simple geo.position META tag, yet some clue to the accuracy may be given by the number of digits in each element. One Minute of Arc of Latitude corresponds to one Nautical Mile (1852 metres, or 1.8km). Latitude Accuracy Mile Km

49 1° 60 111 
49.1 0° 6' 6 11 
49.30 36" 0.6 1 
49.320 3" 0.06 0.1 

In a commercial GPS set, the standard accuracy is less than 15 metres, now that SA has been turned off. (May 2000). DGPS sets may allow accuracy better than this.

Extent
As currently defined, the geo.position tag describes a point, not a region. It is thus unsuitable for describing an extended area. What constitutes a region may depend on the scale of map or geographic search used; for instance, the country of Andorra (geo.position 42.5;1.5) may reasonably be represented by a point on a map of the world but not on a map of Andorra. It is the responsibility of the user to determine if a point representation is meaningful given the intended audience of the tagged document.

Future versions of geotags may incorporate a region element, as do other metadata standards.

Position Datum
In cases where the resource position is less accurate than a few kilometres, datum issues may be ignored. For accurate positions, however, the WGS-84 datum should be used. In North America, this corresponds to the NAD-83 datum. Many maps and charts still use the older NAD-27 datum, and there is a difference in coordinates of as much as a couple of hundred metres, depending on the exact position. Conversion software is available, and most recently issued topographic maps and charts will give the offsets.

Tag Placement
In accordance with the HTML 4.0 specification, META tags (including geo tags) should be placed in the HTML document header, between the <HEAD> and </HEAD> elements, for instance:
<html><head><title>My Document</title>
<meta name="geo.region" content="US-WA">
</head>


TGN etc. Cut/Paste
The Geo Tag Generator allows geographic positions from certain popular sites to be pasted in using the browser clipsheet. Bold text is required. Examples:

TGN: Lat: 49 21 N Long: 123 05 W
Tiger: Scale: 1:218074 (Centered at Lat: 38.89000 Lon: -77.02000)
CPCGN:  Latitude - Longitude : 49? 16' 00" N - 122? 57' 00" W


Check Map
The Geo Tag Generator incorporates an optional check map which may be used for a quick sanity check on geographic co-ordinates. Co-ordinates may also be checked against country and region codes, if given.

Icon
Pages which have included geotags may include a geotag icon to indicate that geographic search capability may be available. To use the icon, copy the image to your system (e.g. right-click "Save Image As") and include the following HTML code:

<a href="http://geotags.com/geo/" target="_top">
<img src="geo2t.png" alt="Geo Tagged for Geographic Discovery"></a>

Referenced from: http://geotags.com/geo/geotags2.html

Notes
The tags describe the position of the resource described on the page, for instance a beach or restaurant, not the company hosting the page, the company managing the resource, or the server hosting the page (cf. RFC 1876)

The tags are described in terms of current HTML practice, which does not preclude them being represented in another manner such as RDF or XML.

The tags are intended for use by a wide base of authors who are probably not well versed in GIS or formal metadata techniques. They are not intended to supplant, and do not address as many issues, as such formal standards as FGDC, GILS, TC211, Dublin Core etc which should be used where applicable.

Hierarchical Position:
     World (facet) 
  ....  Oceania (continent) (P) 
  ........  Australia (nation) (P) 
  ............  Western Australia (state) (P) 
  ................  Perth (inhabited place) (P) 
 

Place Types:
 inhabited place (preferred, C)  ............  settled 1829
city (C)  ............  established 1856
regional capital (C) 
commercial center (C) 
financial center (C) 
transportation center (C) 
university center (C)
 


Unfortunately search engines determine a websites location by the location of its webhost (server).

So, if your focus is a specific locality, then it’s vital that your site is recognised by search engines as being from that territory. Local websites are featured more prominently in local versions of the search engines, there web surfers are given the option to see only pages from their location excluding foreign based sites.

Hosting up until recently tended to be fairly expensive in comparison to overseas in the US. For this reason many Australian based sites found themselves running into problems when they tried to save costs by hosting their sites in the USA.

Also, don’t take it for granted that by hosting with a local Perth based company that their servers will also be based here in Perth! Many Perth hosting companies locate their servers either over East or overseas. When signing up for hosting contracts, if location is an issue for you, always check that the servers are located geographically where you would expect them to be.

There are several factors that are I believe effect the location of a site in terms of search engines:
•             The top level domain extension (.com.au, .fr, .de)
•             The location (IP address) of the website host
•             The geographic location of the domain registrar
•             The language that the site is written in
•             The location of incoming links
•             On page factors (addresses, telephone numbers)
•             Registering with Google Local

Obviously some of these factors hold more importance than others, some I believe are used only marginally, others perhaps not at all, or their effect is too minimal to test.

Using The Correct TLD (Top Level Domain)

This is your best case scenario. You have a site that you’re targeting to local Australian consumers, registering a .com.au site will pretty much guarantee that you’ll be found in the Australian search results, even if you choose to host out of the country. If however you were unable to obtain your business name in the .com.au domain space and went with a .com (maybe because it was cheaper!) or other variations, this may not be the case!
Website Host Location

There are of course many instances of websites that are targeted to a specific country but are using a generic top level domain such as a .com or .net. In these cases simply ensuring that your hosts servers are geographically located in your marketplace should ensure that the site is recognised as being local.

Geographic Location Of The Domain Registrar

I feel this is a factor that is sometimes overlooked by many webmasters, but as Google has access to the geographic location such as the location of the domain registrar it would make sense for them to make use of this as well. This along with other registrar information such as Whois data could well be used as a ‘tie-breaker’ when country-specific TLDs are hosted elsewhere. For example many country specific TLDs such as .fm, .cc, and .tv are now being used because of the brandability of the domain extension. In cases such as these where the TLD extension is indicating one location and the hosting location indicating another then it would be a logical step to make use of the information available from the domain registrar. Tie into this a further step, by ensuring as much of your actual business details (name) and address is accurately reflected in the domain name registration process and thereby reflected in the whois details for your site. This I believe is a factor now being used by Google Local in their lookup for local search.

Site Language

Obviously as well as it making sense to make sure your site is written in the language of the search visitors that you’re looking for, it may well be one of the factors that a search engine may use to help determine the location of a site. It certainly isn’t a defining factor though as it’s relatively easy to find foreign language sites within the AU only search results.

Location Of Incoming Links

As above, the idea is that a search engine can use the location of incoming links to determine the site location. Again, I’m sceptical if this is any more than of marginal importance. I have seen lots of sites with low quality link profiles that consist of largely overseas located links and of course news sites with very few links seem to have little problem get geographically placed before backlinks have had a chance to develop.

Site Addresses/Telephone Numbers

It is simply good practice to have local contact details for local markets. There is also speculation that this may be used to place a sites location. Again this is difficult to test but I’m doubtful if this is would be anything more than of marginal importance. (Consider including local and the international prefix in your phone contacts for example).

Registering With Google Local

Again, hard to test, but it would make sense for Google to make as much use of all the information that was made available to them. Yahoo and MSN also have similar local services.

Monday, September 6, 2010

Robots.txt - informing search engines

The robots.txt file is stored in the root level directory of the website to inform search engines how to interact with the web page, what to, and what not to go into and list in their directory. The following string is the format that is used and provides an example layout.

The file must reside in the root directory of your web. The URL path (web address) of your robots.txt file should look like this: www.yoursite/robots.txt

User-agent: *
Sitemap: http://www.yoursite/sitemap.xml.gz
Disallow: /secure/

To exclude ALL robots from the server:

User-agent: *
Disallow: /

To exclude a single robot from parts of a server:

User-agent: Named Bot
Disallow: /private/
Disallow: /images-saved/

The Robot Tag in Source Code

Where a robots.txt file can’t be uploaded onto a website server, the following robot tags can be included on the individual html pages:
<META NAME="robots" CONTENT="index,follow">

Robots.txt resources:
http://tools.seobook.com/robots-txt/

Robots.txt File Generator
http://tools.seobook.com/robots-txt/generator/

Analyze robots.txt
http://tools.seobook.com/robots-txt/analyzer/

Improving on Robots Exclusion Protocol
http://googlewebmastercentral.blogspot.com/2008/06/improving-on-robots-exclusion-protocol.html

About /robots.txt
http://www.robotstxt.org/robotstxt.html


Tip: the robots.txt file is not the place where you should include comments in the code, as these can sometime be incorrectly misinterpreted and cause problems with the search spider.

Thursday, September 2, 2010

Comparing 301 redirects and the canonical url tag

The Canonical URL tag attribute
<link rel="canonical" href=http://www.yoursite.com/page>
is very similar to the use of a 301 redirect from an SEO perspective.

Effectively you are telling search engines that several pages should be considered as one ,which the 301 does, without actually redirecting visitors to the new URL.

Note however; a 301 redirect redirects all traffic (search bots and human visitors), the Canonical URL tag is just for the search engines to see, meaning you can still separately track visitors to the unique URL versions. As well, a 301 is a better indicator that multiple pages have a single, canonical source. The power of a 301 redirect is in its ability to carry cross-domain functionality, meaning you can redirect a page from one domain to another and carry over those search engine metrics. This is not something that you can do with the Canonical URL tag, which operates exclusively on a single root domain (it will carry over across subfolders and subdomains).