The Three Pillars of SEO
I’ve been doing SEO for a long time – as a hobby for my own sites since the late 90s, professionally since 2004 – and while hypes and tactics come and go, I’ve learned that the core elements of SEO stay the same. When I began teaching SEO about five years ago for universities, institutes, and business clients, I really started evaluating what I knew about SEO and how best to convey that knowledge to my students.
The standard way of teaching SEO, usually as part of a (digital) marketing course, focusses on what I call ‘treating the symptoms’: identifying SEO issues on a site and applying fixes. While this is a decent way to give students a basic understanding of how to do SEO, it doesn’t actually teach them anything about why they need to these things. There’s no deeper understanding of how search engines work, so students lack the theoretical underpinning they need to distinguish hypes and fads from valid long term tactics. When you teach SEO as a collection of ‘treatments’, students are ill-equipped to evaluate new treatments – i.e. new hypes and tacics – to determine whether or not they actually fit with how search engines operate.
The other approach is to teach SEO as an aspect of computer science, specifically Information Retrieval. After all, what search engines do is an advanced form of Information Retrieval, so a good understanding of IR will lend itself well to a prosperous career in SEO. The problem with the IR approach is that for most students it’s far too technical and challenging, and takes years to get a degree in.
I figured there had to be a middle ground – an approach to teaching SEO that isn’t too technical, but that still manages to equip students with the right knowledge of search engine functionality to help them flourish in their own careers without being overwhelmed by the latest fads.
The Three Pillars
- Technology: technical underpinnings of a website: HTML code, URL structure, HTTP status codes, XML sitemaps, etc.
- Relevance: content, title tags & meta descriptions, headlines, content structure, readability, topical focus, etc.
- Authority: links and citations, sending the right signals to search engines that your website is a trusted source.
I’m not the first to have developed the three pillars approach; Dave Naylor gave a talk about his version of the three pillars of SEO back in 2010. Below I’ll outline what each of the three pillars means for me.
A website’s technology is all about making sure search engines can crawl and index all your content and in the right way. This means having clean HTML code, properly structured URLs (with no parameters, if at all possible), the right HTTP status codes for the right kind of webpages (so a Not Found page results in a 404 code, not a 200 or 302), proper XML sitemaps, etc.
Most SEO audits will spend a lot of time on a website’s technology, and the platform a site is built on will have a great deal of impact on how SEO-friendly a website’s technological underpinnings actually are. WordPress sites tend to tick a lot of boxes straight away, while sites built on .NET can often be total SEO nightmares.
When search engines can crawl all your site’s content, that doesn’t mean they know what your content is about. The relevance aspect of SEO looks at the various elements of your content to ensure it can be properly interpreted by search engines.
This means having optimised title tags and headlines, properly structured content, sufficient topical focus on a single page, including the right semantic signals, etc. Writing great SEO-friendly content is now more important than ever. Basic SEO audits will focus on this aspect of your website.
As we all know, great content alone is not enough to rank well in search results. You need to be seen as a trusted source, which is where the authority aspect comes in. With sufficient links from other trusted websites, your site will be seen as trustworthy as well and search engines will rank your content higher in search results for relevant queries.
The link analysis and competitive analysis aspects of SEO apply here, with many different signals being used to determine the authority of a website and thus its rankings for relevant search queries.
Search Engine Processes
Initially I used this Three Pillars framework with an emphasis on the ‘treating the symptoms’ approach, ensuring my students were using the right tools to identify the right issues and applying the right fixes. But I’ve since realised that these three pillars align pretty well with the three main processes that make up an internet search engine.
While search engines are vastly complex pieces of software, at their core they’re made up of three distinct processes. Each process handles a different aspect of web search, and together they combine to provide relevant search results:
- Crawler: this process is the web spider that crawls the web, follows links, and retrieves content.
- Indexer: this process takes the content from the spider and analyses it. It also looks at the links retrieved by the spider and analyses the resulting link graph.
- Query Engine: this is the front-end of the search engine, where search queries are processed, search intent and context interpreted, and results shown according to various ranking factors.
The crawler process of a seach engine discovers new and changed content, as well as all the hyperlinks it can find. A crawler starts with the list of previously discovered URLs, as well as URLs supplied via XML sitemaps (which is why they’re so important).
You can exert some measure of control over how the crawl process treats your website by giving it directives to follow in your robots.txt file. Most search engine crawlers will obey these directives. You could also intercept crawlers based on their IP address and redirect them to where you want them to go.
The content and links found by the crawler are then passed on to the indexing process, which tries to make sense of it all. The indexing process analyses the webpages retrieved by the crawler and assigns relevance values for the various different keywords it finds in the content, which are used for ranking purposes.
It also analyses the links found by the crawler and compiles link graphs, complete with weights and values that can factor in to the ranking of these webpages. Here too some measure of control can be exerted over how an indexer treats your webpages, primarily through the use of robots meta tags and canonicals.
When a user performs a search, the query engine attempts to understand the intent and context of the search and supply the best results. It’ll retrieve the most relevant documents from the index and apply the various ranking factors to present a list of webpages to the user.
Intent and context are important because it allows the query engine to present different types of results – local, news, images, etc – depending on what it believes is the right information for the user at that precise moment. Things like the device being used, location of the user, search history, and query intent all factor in to the results shown by the query engine.
Pillars and Processes
As should be evident by now, my three pillars framework aligns almost perfectly with the (admittedly simplified) outline of the three major search engine processes. As such I believe my three pillars framework serves as a great approach to teaching – and understanding – SEO.
Each of the three pillars aligns with a search engine process. More importantly, I think it helps when a SEO practitioner understands which search engine process they are optimising for. Are you implementing a structural change to the website to make the crawler more efficient or to help the indexer make sense of the content? Is a specific tactic intended to improve link graph weightings for the indexer or to help ensure the query engine sees your webpage as a more relevant result?
Many SEO fixes and tactics will cross between different pillars and processes, but you’ll be surprised to see how many different elements of SEO fall nicely within one of these three main areas.
So as a result I believe I’ve found a great framework to teach SEO. It’ll give students all the necessary information to apply SEO tactics to their websites, but also provides a basic understanding of how search engines operate and how each SEO tactic fits with those processes.
This is of course just my own approach to teaching SEO, and I suspect it has its flaws, as it’s based on my own incomplete understanding of search engine technology. Yet so far I feel it’s resulted in a clearer picture of SEO for both myself and my students.
I’m keen to hear from other SEO professionals and lecturers. Do you think this Three Pillars framework has merit? Would you change it, or do you have a different preferred framework for teaching a proper understanding of SEO? Your thoughts and contributions are welcome in the comments.