Data BlindnessNothing in your search marketing campaign is as important as the data you collect and react on. However, there are basically two types of data that SEOs look at:
- Campaign Data
- Third Party Data
Campaign data is worth its weight in gold, but too many SEOs become blinded by third party data, and allow it to affect decision making about their marketing campaigns. Third party data is valuable as a learning tool and can help create great foundations for testing and segmenting campaign data, but at the end of the day, marketers must leave major decisions to campaign data alone.
The Internal No-Follow Discussion
Ever since SMX Advanced in 2009, when Matt Cutts made his announcement about the internal use of No-Follow changing unbeknownst to SEOs, I have heard continued opinion on the debate.
Conjecture is always good. However, what really struck me as odd was the large number of SEOs that ran out and began making wholesale changes to websites without digging into the data available to them for their own sites. They became so sold on the mathematical concept of link dampening and its relation to this change that they did not look at the websites performance on an granular basis to see how the change would affect their sites. In the end, I heard quite a few stories of SEOs having issues stemming from site changes around this issue. Was “no-follow” the culprit? I don’t know, but I do know if they based their decision on historical data, and setup proper testing to collect more data, they would not have run into these issues.
SEOmoz’s Correlation Factors
SEOMoz took a bunch of heat for putting together a great presentation (delivered by Rand at this years SMX Advanced) and post on the correlation factors in Google and Bing. Too many people spent time questioning the math behind the post, and trying to look at the SEOmoz post as some type of reverse engineering of the algorithm. What Moz was doing was creating a great piece of content for SEOs to see that in the end the one decisively clear metric that matters in search today is links, that and exact match domains are still able to kill it.
SEOs became blinded by data points suggesting that somehow .orgs rank better than .coms, but the data didn’t support that, what it did support is nominal correlations except in the place of links, which showed a true mathematical correlation to ranking.
Why didn’t the people of search simply look at this post as a significant piece of content, and then begin to dissect their own verticals for correlations based on their mathematical standards?
That type of data and analysis could provide critical campaign altering benefit.
Being Blind is Simple
The way information moves on the web it is easy to become completely focused on third party data. As consumers of media we read or view the content, discuss it, and share it. The process replicates itself in some cases, and these are where we become blinded by theory, and lose site of the reality of data at our disposal. Perception outweighs proof.
You cannot allow your inquisitive mind to win over your analytical mind. If you are running your own business online or working for clients, there is too much money at stake to base decisions on anything but relevant data .
How You Should React to Third Party Data
This isn’t to say that third party data is useless. It becomes useful in a balanced marketing approach. Basically you should use third party data as a catalyst for testing and campaign data segmentation to find the answers to your particular marketing issues.
Utilize analytics and testing to make large scale marketing decisions and not information collected elsewhere, because at the end of the day the effects will be hitting your pocket.
Be sure to follow BlueGlass on Twitter feed here. Or subscribe to our RSS feed here.
Dave, this is an extremely important point. To many people ignore data for their own preconceived strategies and tactics. Another problem that I think is equally relevant is folks dependence on a specific data model. It seems that the best SEOs use data and metrics from lots of different sources to influence their thought process.
When I studied politics we actually learned how to manipulate the data from polling campaigns to get it to say what we wanted with out lying. The same is for SEO, because often times the data/metrics are subjective to the methods and models they are collected/reported.
Joe –
Sometimes even testing is difficult. An SEO must take a scientific, marketing, and business based approach to everything. The complexity of the role is why there is only one Aaron Wall, or Dave Naylor, or Greg Boser.
That is very true which is why there are loads of pseudo science SEOs out there.
Dave,
I would take it one step further and disregard any concern regarding link volume all together. Note I did NOT say disregard link volume, just the concern surrounding it. Here is a simple exercise to illustrate this point:
Q: How many links do I have?
Answer 1: Many.
Answer 2: Few.
Point: Would you change your content strategy based on either answer? Not really (hear me out), in both instances you need to produce engaging content that people read and link to. The answer in either case is, produce content and do as much as you can to get or keep links. Don’t slack off because you’re on top or you wont stay there.
Thats why numerical analysis of links doesn’t matter, its noise that has no impact on strategy. The true differentiator is content.
So what should you or your SEO guru worry about? Identify the right market segment, choose the correct keywords for that segment, evaluate them, and build a content strategy for them. A final step is to find partner sites to link with (believe it or not this is easy if you have good content).
You are right on here. We have so many people come to us and say “I should be in a good spot because my competitors have x links and I have x times 2.” This kind of thinking is super flawed on so many levels.
Dave I think sometimes people and clients look at the amount of links a site has and thinks they need to have more, when in fact it is the quality of those links that they do not see as the real benefit. Links are a numbers game, but it is the quality links that many people do not understand as being the most beneficial.
Maybe the reason why people choose not to look at the hard data is they feel overwhelmed or intimidated by what they are looking at in the first place. Yes for large sites it can be a difficult and time consuming task, but just think of the money and time that could have been saved in the first place.
Steve I totally agree, but I also think that people are still not used to having such robust data options at their fingertips. In the past people marketed by what they were told was good practice. Now we can shape our decisions in seconds based on feedback.
Sometimes I wish people would market from within their own testing and results, not everything you read about is the best solution for your niche or business model. Data has been their, but now like you said in so many new forms it is a gold mine for your site information. Inspired me for a new post regarding data forms and how to analyze them, maybe you can just put it up on SEJ. :)
I miss the late night ustream talks from years past, you always put ideas in my head and gave me more work damnit! Never knew my last name, just my old handle of @imaccess