: by Jeff Zwelling, CEO of Convertro :
The season finale of a popular 1960’s drama about advertising executives reminded me of how marketing has changed in the last 50 years. Back then, market research consisted of some focus groups, copywriters pitching copy that was based on a hunch, and splashing ads anyplace that seemed logical. Sales figures were about the only hard data that early marketers had to measure the success of a campaign.
What may be more amazing is how marketing has changed in only the last ten years. Even the changes in the data science in the last three years with the founding of Convertro, online marketing is finally delivering on the promise of producing relevant, rich data for the marketer to work more nimbly and economically. The truth is in the data. This is the kind of information that was desperately needed back when marketers were acting on hunches and instinct. They needed hard data.
Back in 2002, marketers were only getting a taste of the amount of user analytics data web traffic would eventually produce. It was becoming clear that more meta-information was coming with Web 2.0 and Social, but few knew just how much data would be produced by 2012 and beyond. Online data comes from everywhere: blog posts that were read, liked, tweeted, emailed, re-posted, commented on and on and on. That’s all data creation and it’s still growing exponentially.
For example, in just one year people will produce 1.8 zettabytes of online data. Now how do you quantify that?
According to Teradata:
…more data have been created in the last three years than in all past 40,000 years of human history.
According to Mashable:
…one way to put it all into perspective is to hypothetically plug all that data into physical objects we all recognize. That 1.8 zettabytes of data, for example, would require 57.5 billion 32 GB iPads to store…
Now it’s no mystery that the vast majority of marketers got their degree in marketing. If they had happened to have spotted this data proliferation early, they probably minored in computer science, with an emphasis on programming languages and analytics. For future marketers, I strongly urge a minor in computer science, to get the fundamentals of programming languages. Of course it’s never too late to learn. Marketing students who don’t learn at least the fundamentals of programming do so at their own risk.
All this data has the suits -- or jeans -- brimming with expectations. The executive committee doesn’t need a full grasp of the technology to know rich data – when harnessed properly – can reveal new markets and opportunities for the company’s products and services, and maybe even the company’s survival. When it doesn’t deliver, who’s to blame? The constant need to optimize performance based on hard data should be a digital marketer’s best friend. And when it comes time to report up the chain, data comes under direct scrutiny. Getting it right the first time may mean the difference between a fully funded program and crisis of digital faith in the C-suite.
So that ever-growing silo of data is both a blessing and burden for marketing. Without truly understanding its value, an effective campaign is as good as a hunch. Ironically, marketers are now faced with the inverse of a problem that they’ve had through marketing history. Until just a few years ago, when marketers crunched data, they were seeking a needle in a haystack, that bit of absolute truth to build a campaign. Today, with so much data at their disposal, marketers are trying to find a specific needle in a mountain of needles.
But that’s what data scientists do. Data scientists are mathematicians who have been trained with the tools to analyze this vast amount of data through the use of various programming languages and data mining techniques to cull just the right and relevant data from the tap. A data scientist can reverse engineer a conversion to determine the value of the different interactions before the conversion. Our data scientists at Convertro developed a multi-attribution technology to return highly accurate metrics without the need for cookies. Multi-attribution assigns a statistically significant weighted value to each touch point, making it finally possible to understand the actual impact of all of a user’s marketing interactions, from their first touch point all the way to their last visit to a website before converting.
Data scientists at Convertro can now even measure the effectiveness of a television campaign, which provides insight not only about the success of the campaign, but also whether it was more effective within a particular segment of on or offline users.
So you can see that the data scientist won’t displace the marketer. They will be working side by side because each has a complementary skill to make the company succeed. By working together they’ll eventually blur the lines of the two disciplines, because not all marketers suffer from math anxiety. In fact I’ve met quite a few to the contrary. And forget the clichéd data programmer hunched over a PC with empty diet soda cans and bags of chips. That image belongs back in the year 2002 with primitive analytics. Data scientists enjoy their day understanding the numbers, and they may be creators of big data insights by night.