A simple method for object-oriented SEO
I have written about object-oriented search engine optimization previously but it’s an oddball concept for which most people probably feel little comfort. In other words, I put forward an idea that didn’t take off. But rather than consign it to the extremely large pile of discarded Michael Martined ideas, I thought I’d dust it off and give it another shot.
In my previous article I proposed that a Web site could be treated as an object that has SEO-specific properties and wished for tools to access those properties (such as knowing which links are passing a specific anchor text). That was most likely too esoteric and theoretical an approach, even for SEO Theory.
What people have asked for repeatedly is a list of concrete steps to execute. So let’s think of object-oriented SEO as a means of organizing your organic search marketing campaigns. It’s the campaign that is the object in this model, which means that your object is an abstraction but that allows us to deal with very real properties.
Hence, if you want to create a tool that helps you look at your object-oriented SEO campaign it has to use a metaphor (a place-holder) for the object (the SEO campaign). Call it a Campaign Object. You actually use them a great deal more often than might have occurred to you.
Having named our abstract object, we can quickly decide upon a list of useful properties for which we already have a means of connection. For example, a typical SEO campaign must have a list of targeted keywords. If you were documenting your campaign, you would assign the keywords to the campaign keyword property:
campaign.keywords[1] = keyword1
campaign.keywords[2] = keyword2
You would also assign search engines to the campaign search engine property:
campaign.engine[1] = google
campaign.engine[2] = yahoo
You would collect your weekly or monthly rankings in the rankings property:
campaign.ranking[engine,keyword] = organic SERP position
Of course, you could work with other objects, such as keywords and search engines. For a Keyword Object you would want to know the following properties:
keyword.monthly_queries = queries
keyword.cpc_target = desired cpc
keyword.competitors = number of competitive sites
For a Search Engine Object you would want to know:
engine.name = google (so you could assign this object to campaign.engine[])
engine.keywords = number of keywords you’re tracking
engine.referrals = number of referrals you receive from the search engine
engine.kw_referrals[keyword] = number referrals you receive per keyword
So far we’re just using our object to store data, which is pretty boring. What we really want to do is know where we are versus where we want to be. So, for example, we might want a set of goal properties for our campaign object:
campaign.set_goal[engine,keyword] = 1
campaign.return_pct_goal_achieved[engine,keyword]
campaign.return_pct_goal_notachieved[engine,keyword]
Just asking your campaign where it is in achieving its assigned goal would produce some number crunching and it would spit out a value.
All this works well enough if you’re programmer wanting to create a new SEO tool, but what about people who don’t have time to fool with tools? Is there a way to transfer this metaphor to real search engine optimization?
I think you’ll find that you can still track a campaign in a spreadsheet or word document and use these properties without having to create a whole new programming language or tool. You will capture the data from your keyword research and ranking reports, but you need to organize so much more data than just numbers. You need to organize the concepts that the numbers represent.
That is, you want to be able to organize your keywords so that you can handle them all like a list or so that you can get individual data from them. In a spreadsheet program, you assign a different tab for each major concept (campaigns, search engines, keywords). You record data such as goals, rankings by search engine, how many links you’ve built, how many content pages you’ve built, etc. for keywords in the keyword tab/sheet. You can query link to those values from other sheets or tally up the number of keywords per engine or per campaign, etc.
I use fairly complex spreadsheets to track data and by organizing data across separate tabs I ultimately allow myself to create summary pages that at least provide me with a high-level overview of what is going on. You can determine how many of your targets you’re missing with the press of a button. You can decide which search engines are worth optimizing for by comparing rankings/referrals for each engine. An engine that sends few referrals despite consistently high rankings may not be worth optimizing for (or you may need to conduct new research).
You can compare multiple campaigns to each other. Knowing how you executed each campaign is not sufficient. You want to know that your specific tasks have produced the desired results. Suppose you’re managing 10 campaigns where you’ve built directory links for 5 of them but not for the other 5. Suppose you’re doing better with the 5 campaigns that have directory links. Does that indicate that directory submissions may be worth your time?
Quantifying the basic tasks of content creation, link building, keyword placement, etc. may seem like tedious work but it provides you with the means to perform some killer anaylsis — analysis you won’t be able to do with server logs or analytics programs. Tracking click-throughs and referrals is part of the process, but tracking what you do is another part of the process.
Every action we take creates information that can tell you something about what works, what is needed, and what is a waste of time. If you can quantify the data accurately, you’ll improve the efficiency of your SEO tremendously.
And that is the objective of using an object-oriented approach to search engine optimization.
www.seo-theory.com
published @ September 16, 2008