5 Things Marketers Need to Know About Data Science
Data is stressful. It’s this enormously complex and massive presence that underlies and motivates everything we do, demands our attention at every turn, and asks us to blindly defy our intuition to follow a string of numbers.
To think that there’s a burgeoning crowd of scientists – some of today’s greatest minds – who’ve made it their livelihood to make meaning out of complex data – is an enormous relief.
For the sake of a marketing campaign we’re running at Radius, I’ve spent the past six months learning about this livelihood – the growing field of data science, and I’ve gained a distinct understanding of what data science is and isn’t (for marketers).
1. Growing interest in data science stems from the democratization of data.
Take a second to think about all the data you could ever collect about your customers. Basic contact info. Social media activity. Web preferences. Interaction with your brand. Millions of data exist about your customers alone, and you can access and track these data in ways you never could before. Ten years ago, you couldn’t say, “this prospect saw my billboard.” Today, you can say, “this prospect saw my banner ad 4 times, clicked on it after shopping for shoes, visited my website, read a blog article about small business trends, then signed onto Facebook.”
We track a lot of data, and the cost of storing them has decreased so significantly since the 1980s that big data is accessible to everyone.
We’re all sitting on top of massive datasets, and data scientists will help us make sense of all this data.
2. Data science isn’t alchemy.
Applying data science to a dataset doesn’t instantly enrich it with valuable insights. There’s no magic algorithm. No secret sauce. You can’t simply hire a data scientist, set them loose on a dataset, and expect gold.
Expecting a data scientist to pull insights from an unscrubbed dataset is like expecting a sales rep to convert 60%+ leads from an unscrubbed list of event attendees. Data come in all shapes and sizes, and applying any code to a dataset requires a lot of discovering, assessing, shaping, enriching, profiling, and distilling before any analysis happens.
3. Data science is a method, not a job description.
Where marketers are quick to box data science into a topic category, data scientists vehemently disagree on a clear cut definition of their craft.
For our upcoming campaign, Donut Days, we interviewed a collection of well-respected data scientists about their perception of the field. One of them comes from a finance and biophysics background. One of them is an astrophysicist. Others stem from software engineering and statistics training. They’re all data scientists, but on paper, they represent richly diverse histories.
What to put in a data science job description is one of the largest points of contention within the data science community. There’s no formulaic checklist that qualifies someone to be a data scientist; rather, data science is the method a statistician or a programmer or an astrophysicist uses to make sense of big data.
4. Data means nothing until you attach decisions.
A lot of marketers are interested in data science because they have amassed an enormous dataset to which they need to attach ROI. We hear all the time that we need to become data-driven, but most of us have no idea how. Of the millions of data we could track, which ones do we need to stay on top of?
I would love to apply data science to my Google Analytics data to discover a magic formula for the perfect piece of content, but this is impossible until I set parameters for perfect content.
“The challenge to your data science team is not to boldly wade into your data and find something interesting. Not so; efforts must be aligned with business goals.”
Ron Bodkin, Founder & CEO, Think Big Analytics (Forbes, 2014)
Replacing data with insights is about as helpful as telling you the exact profile of an ideal customer; until you know where to find more similar customers or how best to reach them, you’re unlikely to take action. Data science can’t help you get from point A to point B with your data until you use it to drive decisions.
5. Through technology, we can all practice data science.
We’re all drowning in data, and we’ll soon be drowning in analytics. We can track and analyze everything, and thanks to technology powered by data science, we can transform data and analytics into predictions to drive decisions. Data science will be built into the technology that we use to power our marketing efforts.
As we explained in this blog post, you probably don’t need to build a data science team to take advantage of this new trend. Investing in innovative technologies, like those powered by Spark, is a much more manageable goal.
“You personally may never need to upgrade your skills to the level of a professional data scientist. Rather, what you truly need is the ability to tap into continuous, proactive and authoritative insights in all aspects of your life.”
James Kobielus, Big Data Evangelist, IBM (IBM Big Data Hub, 2015)
8 Expert Answers to "What is Predictive Analytics?"
When we commissioned Forrester Consulting to conduct a comprehensive survey of B2B marketing lead...
Determining Your Predictive Marketing Use Case
With over 61% of B2B marketing leaders already implementing solutions, it’s clear that Predictive...
Pepsi vs Coke: Why Marketers Shouldn’t Be Fooled By the Technology 'Taste Test'
When it comes to competition, no two brands have had a more fierce, long-standing head-to-head ba...