Here at SwellPath we’re interested in data, not your simple hey check it out our visitors have increased by 100 this month data, we’re talking the crazy kind of how can we optimize the marketing campaign for the new Death Star kind of data. Now when Google released their new correlation tool on May 25th, the data nerd inside me was stoked. Imagine the crazy correlations I can illustrate could mean absolutely nothing (REMEMBER: 1st rule of statistics: Correlation does not mean causation), or could they?
First, the basics. What does this new tool do exactly?
Search correlations in terms of time trends since 2004:
Ever wonder what correlations in total search popularity with “Otter Pops” over the years? Well, it turn out they are very correlated with other summer item terms, peaking in July and bottoming out around New Years . I know, not exactly the Grand Unified Theory kind of breakthrough, but still gets your mind going on what terms could possibly correlate for our clients search terms over time and season. I found real quickly the tool can lead you to some interesting findings about clients. For instance one client we have, Ariat seems to be in sync with the search term “black jackets”. I don’t know exactly what we’re going to do with this data, but this year’s Black Friday could be huge!
“Search correlations” in terms of US locations:
From what I can tell, it’s based on two search terms along with how many people are searching in those locations. For instance ifyou look up “tornadoes” you will see a lot of correlations to terms with tornadoes all in the Midwest where the horrible tornadoes disaster is recently happened. To be honest, I haven’t been able to figure out a good use for this yet. But evidently “Otter Pops” and “Costco jobs” are huge on the west coast. Also, there seems to be a correlation for guns and fox news in the south…yikes!
Upload your own data to compare:
This is the feature that is most intriguing to me as a data hound, but also the most confusing. I’ll admit that it’s been a bit of time since my last statistics class, so the concept of the Pearson product-moment correlation coefficient took me a bit to get my head wrapped around it and have already started gathering some data together for this upcoming holiday season for our retail ecommerce clients. Hopefully, we’ll be able to show you more in detail later in the year.
The draw tool:
By far the most fun feature of this new application, it allows you to draw your own frequency lines over time and see what was popular in search traffic over Google. Ever wonder what was popular in 2005 and now is making a comeback in 2011? Now you can… in seconds, and evidently it’s … Jordan shoes and wristbands (I can see the business plans already). Now is this tool going to give us any practical data we can use? Probably not, but it definitely will provide you a fun time suck for 15 minutes or so.
I’m not sure what will come from this new tool in the future, but I do know that us data geeks are sure to find new interesting insights into the world around us. Personally, besides the retail correlations, I’m interested to see if there is some PPC keywords we can bid for that never came up in our initial research. I’m sure there are already 5 hedge funds that are incubating their new S&P crusher as we speak. So, if you have any useful ideas or thoughts on how we can use this tool in the marketing measurement industry or just want to proclaim your love for Otter Pops as well, let us know.