A man walks into a a doctor’s office and starts screaming “Help! Help! I have ebola and I’m going to die!”
The doctor bravely comes running and takes the man to an exam room and away from the other people in the waiting room. He keeps his staff away to protect them and, without any regard for his own safety, begin his examination. “Why do you think you have ebola?” the doctor asks.
“Because, Doc, I’m sweating and am short of breath. My muscles are burning and I have a headache. Plus, my stomach is all torn up – I’m on the toilet every few seconds.”
“Okay,” says the doctor. “What makes you think it’s ebola?”
“Well, I got on the internet and typed in my symptoms and the search results told me the symptoms were the same ones as ebola. EBOLA! I’m gonna die!”
“It certain sounds consistent with ebola,” the doctor says, “But I have a few questions. First, did all the symptoms come on at once?”
“No,” says the man.
“What happened first?”
“Well,” says the man, “I woke up this morning and my head hurt and my stomach was sick. Then, after I went to the gym, I started sweating and had shortness of breath and my muscles hurt. I’m freaking out!?
“I see,” says the doctor. “Let me ask, have you been to Africa recently?”
“What about people – have you been in close contact with infected people lately? Did any bleed on to you or were you exposed to open sores of ebola patients?”
“No, c’mon doc, I’m not crazy.”
“I see. Let me ask,” says the doctor, “what did you do last night?”
“I met my friends for dinner and some drinks,” says the man. “Why? Do you think I gave it to them?”
“Where did you go to dinner?”
“A Mexican place. Do you think that’s where I got ebola?”
“No,” said the doctor. “Because you don’t have ebola.”
“What are you talking about,” says the man. “I found my symptoms online and they match up.”
“Right,” said the doctor. “But you don’t have ebola.”
“What do I have?” says the man.
Okay, so I’m not the greatest writer of parable on the planet, but you take my meaning. There’s danger in considering data in a vacuum and losing sight of context.
I love data. As a digital and content strategist, I think it’s the bee’s knees; quite simply, it’s what distinguishes marketing in the digital space from all other kinds of marketing. In traditional advertising, we fall victim to the old logical fallacy ‘Post Hoc, Ergo Propter Hoc’ (after, therefor, because of) where we do a campaign and look for correlative results after the fact. But in digital, we can get a good sense of what’s happening with our marketing in real terms and (nearly) real-time… provided we pay attention to context.
For example: Bounce Rate.
Bounce Rate is a measurement whereby a visit to a piece of content or webpage is distinguished from other visits because the entire visit – from entry to exit – took place on a single page. In other words, someone comes to your site from a link or a search result and leaves without visiting any other pages on the site. We all do it all the time. We follow a link, check things out and leave. Sometimes we do it quickly, sometimes we take our time. And in that lies the distinction.
Bounce Rate comes up a lot – from clients and media partners, vendors, etc… It’s not talked about, it’s snarled, smeared, a scandalous term and a badge of dishonor. It’s an antithesis of success. What’s worse, it’s often looked at in aggregate – meaning it’s not just the bounce rate of a particular page or url, it’s looked at as a whole, the bounce rate of an entire site. It makes sense (a little) because the goal of digital marketing is often to keep people as engaged as possible. Leaving is a bad thing, or is it?
People are going to leave your site. They just are. It’s a fact of life. Paying attention to Bounce Rate in a bubble is not an indication of whether or not the piece of content or site was a success. Instead, you need to ask additional questions – how long did they stay? Did they share the content? Did they comment? Did they leave after five seconds or five minutes?
Just like the guy who looked at a list of symptoms and decided he had ebola, looking at single data points is deceptive in the digital space. If a piece of content has a 30% Bounce Rate, but a 9-second average Time on Page, what’s that saying? Well, unless it is a Vine or a Tweet, chances are it was not fully consumed. However, if a piece of content had a 60% Bounce Rate and a 4-minute Time on Page, what’s that saying? Well, it’s saying that, yes, people are leaving, but they are not necessarily leaving unsatisfied. They have gotten exactly what they wanted or needed out of that content and moved on. Or, maybe you haven’t asked them to do anything else; haven’t recommended other content or asked for some sort of conversion.
The point is this: don’t let the specificity of data fool you into thinking it’s precise. There’s still some interpretation that needs to be done. Think about setting up your analytics approach the same way you might an exercise routine – balance the muscle groups for even distribution of the work. Avoid making big assumptions based upon bubble data points. Try to use the numbers in collaboration (like ingredients in a recipe) to tell you better what’s actually going on.
And, for the love of Dog, stop taking medical advice from the internet.