Nobel-Prize-Worthy Marketing Metrics

interpreting marketing dataSo you don’t have a little love for Chuck Lorre’s sitcom The Big Bang Theory? That’s fine. I don’t really believe you. Even though no show will ever usurp the role of ABC’s Revenge in my heart, I’ve got soft spot for Dr. Sheldon Cooper, the Texas-born theoretical physicist with an IQ of 187. In fact, he’s a little bit like marketing data, and that’s why I love them both so much. Sheldon and website analytics are both brutally honest to the point of disregarding other people’s feelings. Bear with me as I explain the importance of metrics, through the lens of television’s favorite nerd.

Raj: Do you have an opinion about everything?

Sheldon: Yes.

Howard: And you just assume you’re right?

Sheldon: It’s not an assumption.

Lesson #1: Never Make Assumptions

The really great thing about interpreting marketing data is that it allows you to make some pretty educated guesses about what could happen in the future. There are a host of predictive techniques, which range from the very simplistic to the PhD-type of advanced.

By looking at a number of data points about how many people read and share articles I’ve written about inbound marketing analytics in the past, I’m able to make a fairly accurate prediction of how well the content you’re reading will do. However, I’m a lot less accurately able to assume how the dinner I’m making tonight will turn out, because I’ve never made that type of quinoa salad before and have zero data points. Even though Sheldon doesn’t always break down how he comes to his opinions, you can bet your Texas toast that he’s performed some cursory data analysis first, if not more advanced projections.

Sheldon: Bet you didn’t know that I play bongos.

Lesson #2: Be Prepared for Surprising Results, But Have a Healthy Skepticism

Can you imagine Sheldon playing a bongo drum, on rhythm? I can’t. That’s the thing with data. It can be surprising, which is exactly why you need to be aware of this nasty little thing called biases before you get started.

Chad Giddings, VP of Marketing at J. Schmid & Associates writes that bias is more possible than many of us think, and “it can come from a bias in human or constituent agendas…or simply from a bias in the numbers themselves.” While this example is both extreme and silly, if you worked in data analysis at a The Society for Scientists Who Played Bongos, you might have a constituent bias to assume that all characters on the Big Bang Theory play bongos, or that all physicists play bongos. If you’re trying to convince your boss that you need to blog less about lawnmowers and more about lawn care, you could omit a data point that skews results. Being aware of biases is critical before analysis, to ensure that nothing is inaccurate.

Sheldon: You are soft. The world is going to chew you, then spit you out.

Lesson #3: Know Your Data

There’s a lot of really smart people in the world, and many of them hate nothing more than bad data. Once your business blog develops a following, you’ll get called out if your sources aren’t up to par. It’s pretty hard to build thought leadership if you’ve developed the reputation as that one blogger who uses Twitter stats from 2009. It’s just a terrible idea. Giddings recommends the following tactics for ensuring your analysis isn’t full of holes:

  • Know Your Data Sources: You don’t need to publish every piece of data that you collect on your company blog, but it needs to be good. Know when it was published, who was surveyed, and read the footnotes.
  • Know Your Nomenclature: Unfortunately, marketers might have a bit of a disadvantage here. While terms like “PPM” (parts per million) are used universally in chemistry, there are few universal abbreviations in marketing. Check and double-check acronyms, and be sure that you’re citing “gross web visits,” not something else.
  • Know That Not all Data Is Useful: Just because data exists doesn’t mean it’s helpful. Additionally, just because it’s quality data, that doesn’t mean you should cite it in your next blog article. Giddings recommends balancing the quality of data with its ability to help you tell a story.

Sheldon: Two tea bags in one cup? We’re not at a Rave!

Lesson #4: Compare Apples With Apples

I hope Dr. Cooper forgives me for using his quotation as an example of what not to do. It was just too opportune to pass up. You know and I know that herbal tea isn’t usually the beverage of choice at raves, regardless of how many tea bags are used to brew the cup. Whether you’re analyzing a data set that someone else collected or preparing your own numbers for publication, it’s helpful to use other high-quality studies as a point of reference to ensure your numbers aren’t too far off. However, be cautious against comparing apples with bananas, both in your data analysis and content creation. The number for first-time website visitors someone received in 2009 isn’t a good comparison for social media shares today.

Sheldon: 50% of marriages end in divorce, but 100% of sundae bars end in happiness.

Lesson #5: Explain Yourself, Please

Taken out of context, data is about as effective of a communicator as Sheldon is when it comes to talking about emotion. Citing a whole bunch of statistics or dumping some charts into your blog articles won’t do much for you. It should be used to tell a story, convey emotion, or convince your readers of something. You’ve got to back it up with some analysis, so nothing is lost in translation.