People love to talk about the ways that search engines determine their rankings. I always advise our clients to stay away from trying to find "the edge of the algorithm" or any practices that are manipulative because of the risk these tactics carry, but the search engines continue to have many limitations on what they can do, and how they interpret what they see.
Therefore it is prudent for publishers to understand the landscape, and do the right things to make the job of the search engines easier. The SEOmoz ranking factors survey includes a great pie chart showing an estimation of the weighting of the various ranking factors:
However, this picture was reshaped back on February 23 with Google's Panda update. With this update Google added the notions of user engagement and content quality firmly into the mix. This led me to more recently propose a different view of SEO ranking signals:
As you can see in this model, I guessed that the broad category of social engagement and content quality now represents a large 20 percent piece of the Google rankings pie. As I defined it, this piece includes a variety of user engagement signals, such as the way people interact with your site, some form of evaluating the content itself, how your metrics compare to competition on a per search query basis, and more.
I should also note that less is currently understood about the way that Bing is using similar signals, so this discussion is oriented around Google, but Bing is likely doing similar things.
Let's Go One Step Further
When I speak with people about on page SEO, I often refer to it as being required to gain "entrance into the competition":
What I mean by this is that you can't compete for ranking on a specific keyphrase unless your web page provides signals that suggest to the search engine that the page is a good match for the user query. For example, if your page is about Tupperware (for example, look at this page), there is little chance that you can get that page to rank for the term "used cars." It just isn't relevant.
Of course, a few years back there was the notion of Google (or link) bombing, where SEOs ran some experiments to make irrelevant pages rank for various search queries solely through implementing lots of links to a web page using a target keyphrase. Search engines have mostly solved this problem.
So far this is all pretty straightforward, but the notion I'm putting out there today is that from an SEO perspective that on page content is not a ranking factor. It is solely about helping establish what search queries your page might be relevant to.
Here's what an adjusted ranking factors chart might look like if you take this into consideration:
This may be a subtle mental shift but I think it is am important one. If you work with clients, or within an organization, with people who have a limited understanding of SEO, you can often find yourself in discussions where they are unwilling to make adjustments to on page content, because they don't see why they should make those changes. They may not realize that the result of that is that those pages end up having no possibility of ranking for a particular term.
There is also the flip side of overemphasis. I have encountered countless people who think that SEO begins and ends with on page SEO.
"I've optimized the site itself, so I'm done, right," they ask.
Well, no, you aren't. It simply buys you a ticket to the competition (it makes you "relevant" to the query).
This is an essential step to success. You don't get to play without it, but there is far more work to be done before you can declare victory. This is the link building, and engagement optimization which make up a full 96 percent of the rankings picture in my adapted chart.
One side note: the only way on-page optimization can enter into the rankings chart is if you engage in keyword stuffing of any kind. This is still something that you can encounter on the web, and I believe that any sort of abusive practices can become a negative ranking factor, but for purposes of my chart, I have chosen to start with the assumption that this type of practice isn't a consideration.
I use this mental model when I think about SEO work, and it helps me get to a really clear picture on how I spend my time. It also helps me get others focused on where time is spent, and how the major components of the SEO world fit together.
You can't participate without addressing on-page SEO (after all you won't be relevant), but you can't win without addressing the promotion and user engagement pieces of the puzzle either.
As always, bear in mind that your publishing strategy should begin and end with understanding what unique value you can bring to visitors to your site and how best to provide that value to them. This is the first and most critical element of SEO.
Being mindful of the limitations of search engines and making their job easier is another necessary step in the process. To that end, provide clear keyword based signals as to what your pages are about. Make sure you are in the competition as your first step, and then set out to win it with an effective marketing strategy for your site.