We could argue until the cows came home as which 2 main elements of SEO were the core components. I’ll cut to the chase and ask you to consider content and links alone this time round. They are (arguably) the most important two variables present in the SEO diagram at this time. It makes sense that the text on the page and the images and so on that populate this page would have some kind of impact on its quality assessment and also it’s competitiveness in terms of ranking against other documents. Links are of interest because the web looks like a web because of all the nodes link together. It makes sense that inbound and outbound ones should have some impact because they provide further information about the page. The question remains however: where is the balance? How much are links worth as far as rankings are concerned in relation to content on the page?
“Content is king”:
It’s really important for the most significant part of the machine: the user. That alone should be enough for us to chime in with “content is king”. Is it as far as rankings are concerned though? If you decided to try and rank a page just for the sport of it, would this be where the core of your efforts would be?
“Good content will get linked to” says Samir and I couldn’t agree more, but then I do see amazing content hidden away because it wasn’t linked to other content properly. I have seen this particularly in academic sites because I regularly sift through a good pile of those during my research. Why did it take so long to find some of those pages? The content is awesome. SEO makes sure that pages get found. How does that happen? Through links. These pages had one inbound link at most.
Can a search engine appreciate good content?
That’s an interesting question. It is undeniable that we need to create quality content for our users. In natural language processing research we strive to get machines to understand language. Since the 50′s research has been ongoing in natural language understanding and after all this time we still can’t really claim to have made a machine that “understands” language as such imho. We have created systems able to process symbols correctly and able to determine the correct output in a closed domain quite well. Natural language understanding requires 2 main ingredients: world knowledge and language awareness.
Which is the most important sentence to you?
“I don’t want to live. I want to love first and then I want to live incidentally.” (Zelda Fitzgerald)
“Only a life lived for others is a life worthwhile” (Albert Einstein)
Both require you to have a certain appreciate for what human life entails, what love is, and the fact that it is a big philosophical question too. We might appreciate Einsteins quote because coming from a scientist his view much like Mother Theresa’s. Zelda had a rather famous author for a husband, perhaps we know they didn’t quite get on seeing as she said “Mr Fitzgerald – I believe that is how he spells his name – seems to believe that plagiarism begins at home”. So in order of relevance which quote says the most about life?
The above just serves to illustrate that it’s subjective, relies on a lot more variables than are available on the page. The semantic web also involves the links between the concepts and how things relate to one another.
A search engine has the role of returning relevant documents to a user. It is a bit of a hammer and nail approach and most things work well enough. And that’s enough. In other applications that use machine translation for example the approach needs to be more accurate. Good enough isn’t good enough. A search engine needs to know what the page is about and how it relates to others as far as the content is concerned. There may be a measure of quality insofar as misspellings are concerned and inaccurate grammar for example but can the search engine understand the language used?
If I have a site full of sentences à la Chomsky like:
A customer retires a trend next to a vanished basket.
A courier trails next to the newest basket.
The magazine gasps on top of the scenery!
Will my page more likely rank for “nonsensical sentences” or for “basket” for example?
Can gibberish rank?
If your link structure is good enough, I don’t see why not. The method is a statistical one after all so as long as you tick all the right boxes the maths should add up right? Until search engines use all the content and everything it has to offer it will always be king for the user but for the engines perhaps there is more value in links. The issue is that the engine can’t easily recognize the gibberish the way we can. If MIT’s work is anything to go by, some humans aren’t very good at it either!
Links count too:
Imagine that you are all powerful and that you own 100,000′s high authority websites. In fact in your chosen SERPS your sites show up for most positions. You can interlink them all and create exactly the kind of structure that search engines covet. Your websites, as long as they didn’t break any “rules” would probably do quite well, despite gibberish sentences too. PageRank has morphed greatly since its beginnings but the main weakness of the strategy itself is the reliance on links and not content for example. If relying on links worked there would be fewer spam issues in the SERPS. It’s not a big secret that we need to expand the ranking criteria, base it more on content than that, and particularly on the language, the context, and the meaning.
So should I just have random gibberish on my sites?
There are a number of reasons why you’d do that, one being to bury a site giving you bad press or a site you just don’t like or something without needs to go into all the effort of writing content for it. Thing is it’s a matter of time before it gets fished out for being rubbish and so it’s a short-term strategy. You certainly wouldn’t want to have any users go there.
It’s like this:
Gibberish text is like going without water.
If you want your site to thrive it’s going to need food (links) and water (content).