seoSearch Engine Optimisation has seen exponential growth over the last year and the good times look set to continue as firms make the most of their web assets during the credit crunch. Google’s webmaster group published their guidelines covering the basics of promoting HTML content - the rest is down to effort, discipline, time and creativity - but search technology is seeing some big changes on the horizon.
The emergence of the semantic web is going shake things up big time and require some old SEOs to learn new tricks. Whilst text indexing will remain the bedrock of internet search technology, e-commerce sites that leverage semantic search technologies will find it much easier to get closer to thier customers. SEOs working for such forward-thinking sites are going to have to brush up on their technical skills and go back to the roots of Information Retrieval.
Why?
The semantic web is about presenting data to search engines. Let’s call these snippets of data facts (their technical name is triples, but facts is a great term to help your understanding).
I can write a web page using current HTML technology and the only facts that I give to the search engine crawlers are:
- “There is a collection of words at this URL”
- “The words in my heading are more important than the words in the body”
- “The document is written in British English”
I could add microformats to the page and include these facts:
- “The document was written by Andy”
- “The document was published on 18th Feb 2009”
Great, I could ask a search engine for all documents written by Andy between July and August 2008, but few people ever will. They ask for short jumbles of words (“cheap holiday Heathrow spain”) because that works well with text indexing technologies available now. That query really means “I want to see details of cheap holidays to Spain that fly from Heathrow”.
Now consider optimising a page to include these facts:
- “This document describes a holiday”
- “The holiday is in Alfarnate, Costa del Sol, Spain”
- “The holiday includes flights from Heathrow Airport, 234 Bath Road, Hayes, Middlesex, UB3 5AP, United Kingdom”
- “The flights are with the Jet2 airline”
Are you SEOs reading this getting excited/worried yet?
How?
As an SSO, it’s your job to get your document considered as an answer to as many questions as possible. To achieve this, there are parallels with current, textual search engines:
- Get your data discovered and crawled by the semantic search engines
- Build trust in your data
- Ensure your data is valid and up-to-date
The big deal here is trust.
Semantic web tools start out with a corpus that they trust, often Wikipedia, scientific or government data. They don’t want to pollute that with dodgy data that’s been hacked around to get better rankings so you’re going to need your whitest hat. SEOs are familiar with Google’s TrustRank algorithm, and you can bet that this will be applied to RDF data too.
Links
As always, links will be hugely important.
A shoe e-tailer could link from his product offering to the manufacturer’s product description data. Affiliates and price comparison sites can link to the e-tailer’s product data. This already happens with HTML pages but it will happen with data objects too.
Guidelines for semantic link building will follow similar themes to those for HTML link building:
- Create compelling content = create useful data
- Build relationships, don’t buy them
- Become an authority in your niche
- Link out to other datasets
- Follow standards = use an existing ontology
- Blog and ping = update and ping
Directories Make A Come-Back?
Although discredited for HTML link building, I think that directories will regain favour as people look for good data - at least in the short to medium term. Directories of semantic data sources will be able to capitalise on the machine-readability of RDF and implement a form of automated quality scoring.
When?
You can start now - even the humblest WordPress blog can add hAtom semantic tags by following these instructions.
More complex sites, such as e-tailers should begin planning to deliver semantic data sooner rather than later - perhaps by integrating it with your affiliate programme. Being the first in your niche will count for a lot when other sites have invest effort in linking to your data.
I’d like to bring your attention to an excellent slide show by Jonathan Mendez that discusses the use of semantic data in landing pages. Lots to think about there.
Are you an SEO? Have your thought how “Web 3.0” will change your working practices? Have any customers asked about semantic technologies? Please leave a comment.
Creative Commons licensed photo by lecasio.
del.icio.us
Furl
Google Bookmarks
reddit
Simpy
Sphinn
StumbleUpon
Yahoo MyWeb
Post On Fire
Google Buzz
Great Post
First of all nice post, short and to the point but informative none the less, secondly thanks for the link to the Wordpress tutorial, that's pretty good too.
As an SEM I am recently exploring semantic web a lot more (along with a bunch of the IR stuff) and this post is just a fantastic summary of why people in the industry should be paying attention.
I think it's obvious this will and to a certain extent, already is affecting SEO/SEM and i'm actually really looking forward to how it develops and how the web community reacts. Currently we havn't had any clients specifically ask us about semantic web but I think as the Web 3.0 buzzword gets passed around more it will definately be coming up in meetings.
As far as the way we work we are already changing the things we do, trying to stay on top of the semantic web standards and making the newer projects we work on compatible at build, not always easy in these early stages though!