I’ve been talking to a lot of people recently that seem to think that Search Engine Optimization is bullshit. There seems to be a rising opinion that it is all snake oil and if people just ignored the search engines then the world would be a better place. This is not true.
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
- 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.
- 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.
- You use a template for each page that results in the search engines thinking that each page is a duplicate of another page.
- 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.
- 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.
I hope this goes someway to showing web owners that SEO isn’t overly technical, isn’t magic and is actually quite common sense. People who know very little tend to republish things they’ve read – link Keywords and H1 tags and so on – because its easy and they think thats what works and what people want to read. Because a lot of people secretly believe there is a secret, special tag that can make you #1 for everything!
A real SEO will not talk about link building, keyword stuffing etc. In fact, Google with their blog and Matt Cutts (Search Quality Engineering) give you all the facts – write content for people, forget the keywords/tips/tricks. They dont read meta-tags that are hidden (e.g. “keywords”).
Develop a site for people, make sure your site is as well connected as you are (suppliers, colleagues, customers, associates, industry bodies) and grow your online presence – engage the community, showcase your opinions and expertise, publish them so people can link, share, discuss or even argue!
@LeoFogarty Pointed me to http://www.seomoz.org/blog/terrible-advice-do-seo-for-users-not-engines which is worth a read.
“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”
I would have to disagree with this statement although I do agree with your overall message which you summarised nicely in your last paragraph.
In the pure sense of things SEO efforts have nothing to do with how well google works.
Simply put, if you create content, by default the words most relevant to the page will be in the text. google will pick up on this and consider it relevant.
If its a widely linked to page then google will rank it higher in the search results all other things equal.
Don’t get me wrong, you have to put a lot of effort into creating content and in the right way so you maximize your chances of being found. But it comes down to two simple things. Content and popularity.
I agree Derek – simply having the right content without having the authority to Rank is crucial. Just having relevance doesn’t rank you – as in just be relevant to a voting constituency doesn’t get you elected – you have to have the votes – and we know from Google that relevant sites linking vote (through their authority)
Hi Derek, David, thanks for the comments. As Caelen mentions, link building is really a PR or marketing function. You can have amazing content but if no one knows about it, no one will ever link to it. Where SEO does come into link building is in the link text and link title text.
I’d suggest that 10 links, even from relevant sites, with the link text “click here” are worse “votes” for the authority of a page than 10 links from equally relevant sites with some kind of relevant text in the link. For example, a link in to this page on our site:
http://www.revahealth.com/dentists/ireland/county-dublin
that uses the link text “Dentists in Dublin” or some similar variation is worth a lot more than a link that uses “click here” as the text.
I would also see SEO itself as making the relevance of the page as clear as possible, both to the end user and the search engines.
Hi Derek
Thanks for stopping by and while agree with you that it’s really about the content. My point is that on large scale sites or site that are not easily lexically analyzed you need to go beyond that. For example a major project that I worked on 5 years ago (www.foneblog.ie) simply could not be crawled by the search engines. It wouldn’t matter what content went up there it just wasn’t going to get indexed.
It takes a major SEO effort just to make a larger site efficiently crawlable. The likes of wordpress and other CRM packages have put a massive effort into this and for those that produce blogs on the back of them benefit from this.
A website isn’t just about original written content. We need to think about all kinds of things that search engines just can’t understand – pictures, video, music, the spoken word. All of these things contribute to a humans understanding of a page. Lexical analysis of a page simply can’t manage this and most spiders don’t even bother access the content.
I was recently at a website of a dublin city center restaurant produced by a well known Irish web application developer that didn’t include the world ‘restaurant’ on the page.
Why?
Because there was a bloody big picture of dinning room and loads of small pictures of plates of great looking food then there was the massive menu. Any visitor on the site instantly knew it was a restaurant . The use of pictures conveyed the meaning very effectively – overall it was great human communication. However for a user now looking for a restaurant on that exact street in Google it didn’t exist.
Now that’s fine if you don’t want the search engines to send you any traffic from people looking for a restaurant where you are. But if you do want relevant traffic from the search engines then you need play their game and tell them what your page is about.
The corollary of this is also true. Frequently search engines can send lots of irrelevant traffic. In our case, on our dental pages we referenced the educational requirement of medical professionals. This resulted in a lot of traffic for people looking for medical courses. This serves no one, not us, not the search engine and most of all not the user.
Now why wouldn’t we want free additional traffic even if its not relevant, after all bandwidth isn’t that expensive. The reason it that is ruins your metrics and introduces noise into your KPIs. It also resulted in irrelevant contacts for our clinics which reflected badly on us. Good SEO identifies and rectifies these issues.
Google recognizes this and have created a suite of tools to diagnose issues with SEO (web master tools). Google also expends a huge amount of effort in providing guidelines. The volume of these guidelines is massive and constantly changing. The reason they provide these guidelines is help publishers shape their content so that they can better understand it.
A lot of SEO has become synonymous with good application or website design. For example on your website you have a sitemap, you use a very SEO’d title, a meta description so the search engines will have something meaningful to display to users. While you may not have consiously have been thinking you were SEO’ing your site, these elements serve no other purpose.
For bloggers and other small scale written content publishers I would agree entirely with your sentiment. They can sit on the SEO shoulders of their CMS package and only worry about the content they create from a user’s perspective.
I wasn’t disputing the importance of SEO efforts and maximizing what you can for your website. Everyone should maximize for their own benefit.
I was just pointing out that Google wasn’t built with the hope that SEO would be used on all websites. It only hopes that people would use best web publishing practices e.g. use of correct html tags and in a natural way. To me this isn’t SEO it’s just doing it correctly in the first place. Again I’m only disputing what google’s point of view is. I agree completely with the stuff you suggest for a website owner.
I think the people who knock SEO are the ones that either don’t understand how the ingredients come together or simply can’t get a website ranking. Website don’t just appear with generous rankings. There has to be some sort of catalyst to push those websites and allow them to climb. SEO is not just SEO either, it is proactive online marketing. PR, articles social media are all forms of SEO and if a person chooses not to use those forms of marketing they will remain invisible.
Top tip for today: We were talking to a guy about SEO for his site today and the navigation on his site used javascript to create the list of links for each section. With the JS turned off, the links disappeared, so Google wasn’t seeing or following them, and so wasn’t passing pagerank to the sub pages from the homepage. So, remember to check how your site behaves with JS turned off!
[...] search engine blog in an online marketing list, but it’s one of the best places to read about SEO, Duplicate Content or revenue [...]
@OP – I think the last comment is blatant Blog Spam – ironic given the nature of the post and the link that “Joe Smith” has placed there.
@Derek – the very workings of Google (Ranking based on Authority where site is relevant) is at the very heart of SEO. SEO is about developing relevant content and growing the authority. You need the two – relevancy is not authority and doesn’t bestow it. You can’t vote for yourself.
Maybe the conversation is – what is responsible SEO and what isn’t?
Joe Smith has been consigned to the Spam folder David. I’m interested in hearing any thoughts you have on responsible SEO, particularly anything you think might have changed from being acceptable to unacceptable, or vice versa.
@PhilipBoyle
The spam artists are getting a lot cleverer – I’ve even noticed Irish SEO’s posting a lot on my own very simple blog.
I think SEO exists as a real, recognised function of a website. All of the search engines have an SEO forum, advice page – and most of them rank quite highly.
If you have a website and sell BMW’s in Dublin and you develop rich, interesting content, you structure your site cleanly, you don’t overkill meta-information and you read what the search engines direct you to do – for example, using the Description tag to convert people from a search page and not to “trick/fool/other” the search ranking algorithm, then this is ethical. You are keeping the relevance of your content to what you do and what your provide. That way if someone is looking for “BMW Clare” – then you aren’t going to appear there etc just to get extra traffic (albeit with a high bounce rate)
I think SEO includes things like : site architecture, inter-site connectivity – and now more than ever before – conversions. Which means narrowing visitors back to low bounce figures and designing the flow of the site to generate as many pageviews and conversions/enquiries as possible.
I think spam and badly generated content, such as repeated keywords/variatinons, nonsense reviews of products, listing products that aren’t stocked in order to bring in traffic to sell other products and anything that doesn’t give the user what they want is unethical to a degree.
That’s probably where personalised search comes in (which is now available even when you aren’t logged in, by default) – it will allow you to block sites like these going forward. You can also a message using Search Wiki and the new Side Wiki tool on the latest Google Toolbar (firefox & IE) – which could be a warning of sorts to some and high praise on others…
@David It’s going to be very interesting to see how personalised search pans out, especially given that the vast majority of users won’t know it is happening to them. It will also have some ramifications for SEO – if a person has a search history (and most will) it will be harder for new websites to break into their personalised search results as the person’s commonly visited sites will be favoured. I wonder how well Search Wiki and Side Wiki will catch on – I certainly haven’t been using them myself.
@Philip
Funny thing about technology is how even the biggest guys fail to learn from each other. twitter is struggling with spam although its doing a great job. Skype struggled with it too. The funniest thing is that Yahoo had far more sophisticated technologies and a larger user base than both twitter and Skype have now – and they had it ten years ago.
Now Google has an open season on leaving comments beside websites which will be publicly visible to all and sundry when logged into Google but they have no verification system/anti-spam system for it…
@david
This is so true for all areas of software. How many times have engineers written code to send and receive emails? 10s of thousands I’ll wager and nearly every time they write it from first principles. Software is something that is inherently reusable, we should be able to create a body of knowledge so that we effectively re-use tech and don’t have to re-invent it each time.
@Caelen – at the risk of swerving 90 degree’s off topic – would you not kill innovation?
@david – true if we abandon all primary research, however I’m just talking about learning from other. If your going to write a email gateway then look at state of the art implementation and learn what you can first. If you are going to launch a platform that can be spammed then look at other platforms that can be spammed. Really I think I’m just saying what you were saying.
@Caelen No point re-inventing the same wheel and there certainly is a lot of re-usable code reproduced on the web.
As an aside, while a good point you make, an unfortunate example that you picked the e-mail gateway, to wit the HELO standard developed probably 30/40 years ago is directly responsible to the huge spam problems we have now as the security-less prototype became the Model-T standard immediately afterward. Definitely due for re-development but too widespread now?