I love strategy. I’ve loved it since I was a kid and spent hours playing chess on Yahoo! Chess – not to mention late-night online AOE battles. This love of strategy is one of the things that attracted me to business, and more specifically, marketing. I’ve also always been a bit of a computer geek and learned the basics of programming, hardware, and web development in high school. One of my first jobs was an IT support role at my high school. I know, I’m pretty cool…
Imagine how excited I was when I first saw Google Analytics and had someone talk to me about SEO, PPC, and digital marketing in general about 6 years ago. The light bulbs went on and I was hooked. It quickly became clear these disciplines were perfect for those who love strategy, marketing, and have a technical bent as well.
About a year ago when I moved into my current role as the Director of Link Strategy I decided I needed wanted (ok, yes, needed) to master the art of link strategy development. Fortunately, I was encouraged to pursue this interest even though it didn’t immediately provide a ton of value. The product however has become a staple of how we've handled link development as an agency for the past year. I’m really excited to get to share some deep insight into the approach and hear feedback from those who love strategy, links, marketing, SEO, and all those other related disciplines.
In order to understand the approach (and some of it's limitations) fully, you’ll need to keep the following things in mind.
- At an agency with teams serving dozens of SEO clients at any given time (and growing), I'm just a single person. These clients range from mid-size businesses with little known brands to major Fortune 100 companies (Read – the strategy had/has to scale in many ways).
- I’m not a mathametician/statistician. There are undoubtedly ways to refine this process and gain even more insights if you have a strong grasp of statistics and other advanced mathematical disciplines. This approach however has proven accurate and insightful over dozens of iterations.
The process – on to the nuts and bolts:
Note: Some details are omitted that might have a bit of secret RKG sauce in them (doesn’t that sound good?), but this will give you the main points of what we walk through and why.
Here’s an overview of the process:
Step 1: Identify your true SEO competition
I’m still a bit surprised when we talk to clients about their competition. About 50% of the time, who they identify as their competition in search isn’t even close to the truth. If they’ve never engaged in SEO, no surprise, but that often isn’t the case.
The first thing you have to do is make sure you know with whom you are really competing. Without this, you don’t know if you’re doing great, getting killed, or most likely, are somewhere in the middle from a backlink perspective.
In order to accomplish this, we utilize a set of keywords and build out an SOV (Share of Voice) report. Those showing up with top SOV percentages across our chosen set of keywords are marked as competitors. No if’s, and’s, or but’s about it. These are the people that are showing up the most in the SERPs for the words we care about. These are the people that must be beat on a consistent basis.
Step 2: Building the comparison
With identified competitors in hand, it’s time to do some comparisons. Now, there are a lot of ways to do this that are much more in-depth, provide great insights, and allow for extreme fine-tuning of a strategy. They also tend to take quite awhile, and the conclusions can be somewhat tricky to explain to a CMO who has no idea what anchor text is. This practice has its place, but it’s not here (more on that later). Bonus: If you find someone who is able to handle this level of analysis and turn it into practical, actionable, business speak, they are pretty valuable - hold on to them.
In order to draw valuable and actionable insights without going to extreme depth, we use the metrics provided by Open Site Explorer - specifically MozRank, MozTrust and Domain Authority. I’ll also note that Domain Authority is a key concept to understand in depth.
If your knowledge is loose on what things are packed into these metrics, I suggest studying them more in depth. The true power of this method lies in what can be extrapolated when these metrics are cross analyzed.
After you’ve identified your competition, gather the MR, MT, and DA for each (BTW, Excel works really well for what we’re doing…). *Note: Whether you use subdomain or domain level MR and MT is your choice and gets into how separate you think Google views subdomains from each other.
After you’ve pulled the metrics (and make sure you do it for your own site as well), take a look at the competitors and see how they are stacking up. There is a lot of information present in just these few things that can give some insights into potential problems/weaknesses the competition might have, or that you might have. If your site is killing it from a backlink perspective relative to the competition, but your SOV sucks, that’s worth thinking about. There are many possibilities, but at the least it should alert you to a potential problem.
Step 3: Analyzing what you’re seeing (the chart)
This is where the magic really happens. Before we dive in, here’s some things we know:
1) MozRank is made up of quantity of links as well as quality of links (includes MozTrust within it from what we theorize). It also includes some juice from internal linking.
2) MozTrust is a score measuring the quality of the backlinks going into a page, subdomain, or root domain. It does not include any scoring related to quantity.
3) Each of these is based on a logarithmic scale. The base log is a bit of a mystery, but we assume it’s around 10.
Essentially, we’re going to take the data gathered and throw it into a matrix that will show us things like:
- Whether or not your site is competitive against the competition from a backlink perspective when compared by MR and MT.
- In general, how badly you are losing to, or beating, competitors.
- Whether you might be able to get away with more “quantity” and less “quality”, or whether you really need to focus on quality over quantity.
- When you divide the competition up based on a midpoint, whether you are more like the top dogs or the bottom of the barrel scrapers.
At this point, it’s important to calculate your midpoint. We use a formula that calculates the midpoint of a set of data points set to a logarithmic scale. I don’t share it here, but it’s relatively easy to come up with some methodology on this by browsing online.
After the midpoint has been calculated, you can build the graph. Ours usually ends up looking something like this:
The strategy plot clearly divides the competition into four quadrants. In these quadrants is the key. We setup our plots with MR always on the Y-axis and MT always on the x-axis. With a little thinking you can start to see the insights that can be pulled from this comparison. It’s also visual which is great for clients and team members.
This plot drives the tactical recommendations made. At this point, I’d like to exercise a pet peeve of mine – blogger outreach, forum link building, etc., are all tactics that should be used depending on what your strategy calls for, they are not really strategies on their own.
Most of the time, there are a few situations that a site find itself in:
- You have lots of backlinks, lots of good backlinks, lots of linking root domains and all the other good things you need. Essentially, you have the power, you just need to utilize it correctly.
- You are lagging the competitors you really need to beat from a backlink perspective. This means you need to come up with efficient ways to improve your backlink profile so that you can move into situation 1. How do you do this? There is a simple equation I like to use:
Content + Promotion = Links
Looks simple, but trust me, it’s not (as many of you know). What goes into each of these elements so that you end up with the correct result is a topic for another (or several) post.
Wrapping it up
If you use a method similar to what I’ve laid out above, it’s important to consider all the different data points (and others you might have), and not simply go off of what the chart is showing. It's a big piece that provides a lot of value, but your insights become even better when you start to ask more questions based off of what you’re seeing.
- Their backlink profile stinks, but their SOV is high. Why?
- What types of keywords are primarily making up our seed list (head, long-tail, etc.)
- Based on what we’re seeing, what options do they have now and what can we hope to do in the future (what’s realistic to accomplish).
Things to keep in mind
In order to effectively act on what you’re seeing and achieve your goals for SEO (and specifically backlink development), it’s very important to think through the following items:
- What resources are available to help with link development and how much of each?
- What’s reasonable to accomplish and in what time frame? How does this tie into our over-all SEO goals?
- Remember, we’re only seeing a part of the picture. Our understanding is that SEOmoz indexes around 10-20% of the web. Likely the most important parts, but still, it’s limited.
A final thought on SERP analysis
Search is maturing and there are continual tweaks, removals, and additions to how search engines digest the web and organize it, and I'm not talking about page titles. Link development is still important, but I believe it’s not the end all. Deep SERP analysis can not only help you identify opportunities to improve your linking, but also give you insights into why certain things might be ranking beyond just links. Personally, I believe it’s possible to hit your head on a ceiling from a backlink perspective (rankings stop going up).
Spending some time analyzing the SERPs for any specific target pages is always a good idea as you may find a situation where backlinks don’t really seem to be that crucial, or perhaps there really seem to be a few types/situations that are more important to engage with than others. Keep your eyes open, this method for identifying a strategy is only a starting point.
Here are some things to get you thinking about the future and maybe even the present more than we realize: