Big data is an umbrella term. It encompasses everything from digital data to health data (including your DNA and genome) to the data collected from years and years of paperwork issued and filed by the government. And that’s just what it officially covers.
Like a Russian nesting doll, big data also houses legitimate ideas and terms that use big data as a source of information, but differentiate from it based on segmentation. The most popular of these nesting terms, if you will, are smart data, identity data and people data.
As of yet, you won’t find an easy Wikipedia page explanation for any of these ideas. Instead, you’ll see them thrown out on Twitter, within the copy on start-up websites and debated by executives — many of whom have yet to fully understand who needs big data, what kind, when and why.
And the confusion is legitimate. Few fully understand what big data is, much less what the term’s offshoots entail. But big data is evolving and smart data, identity data and people data are here to stay. Think of them as the human discovery of fire, the wheel and wheat. Just as those inventions couldn’t have occurred without humans, these subset terms couldn’t exist without big data.
Of course, the human race wouldn’t be where it is today without those three key finds, and these data segments will prove to do the same for big data — making it comprehensible for the masses.
Are these definitions all-inclusive? No, but they will help you to wrap your head around the terms that will influence digital media careers and online experiences for years to come.
Smart Data
Perhaps the simplest to understand of the three terms, smart data differentiates from big data in that it is actionable.
Big data gets its name from the massive amounts of zeros and ones collected in a single year, month, day — even an hour. This type of data is spit out onto spreadsheets, often requiring someone with a PhD in some type of data deciphering to find commonalities, create algorithms, implement those algorithms and then perhaps see a business result.
Smart data makes the PhD unnecessary — though smart data platforms are generally created by data scientists, many of whom hold a PhD. Within a smart data platform, big data sets are siloed, segmented and then visualized according to your business need — and these platforms can be customized for individual teams within an organization.
From the platform, an editor might be able to see that 30% of the people who read the breaking news section on their site on a daily basis also “like” both President Obama and Jack Daniels on Facebook. From here, editors can easily create an email campaign targeting these users. On the marketing side, employees can promote content targeting potential loyal readers that match that audience. And for the ad sales team, they can boost their CPC rates for brands looking to gain traction with Obama and Jack Daniels fans, or can easily print out an RFP response to hand over to Jack Daniels — showing them why sponsoring content on the site will be beneficial for them, the publisher and the reader.
Smart data platforms generally report in real-time and can use data points from multiple data sets (including offline data, spreadsheets, logs, social media, on-site metrics, etc).
Identity Data
Identity data isn’t easy to explain, but think of it as the future of big data. Better yet, think of it as the wheel: the invention that will get us through the next few generations.
Identity data is the force behind predictive modeling and machine learning. It is also the side of data that is most concerned with security. When Target’s data was breached, it was the loss of identity data (or, the placement of it into the wrong hands) that became the biggest issue. Credit card numbers associated with names and physical addresses, as well as email addresses, were stolen by hackers.
These data points are what constitutes your identity data, and combined with your social media data, purchasing habits, on-site behavior analytics and even wearable tech data — your identity data tells the story of who you are in the digital age, including what you like, what you buy, your lifestyle choices and at what time or intervals all of this occurs.
Sound scary? Sound like a privacy issue? Not necessarily. Identity data helps organizations better understand what to keep on their shelves, how many emails they should (or shouldn’t) be sending you and in all, identity data helps organizations with predictive modeling and machine learning — so that before you get annoyed with a company’s marketing tactics, that company’s efforts disappear from your digital life.
Identity data is what the newly hired data scientist over at The New York Times is sorting through. His job is to figure out when you will unsubscribe to emails, at what point in an article you stop reading and who is likely to renew their subscription to the 10 articles you’re allowed to read a week — and why.
The goal? To make the Internet less annoying and frustrating, and to make your on and offline lives seamlessly blend. It’s the future our children will know as second nature. Currently, we are all just trying to figure out the security and privacy issues this type of data collection entails. And we’ll get there.
People Data
You guessed it, people data is akin to the discovery of wheat. No, not every civilization cultivated wheat — but those that did grew exponentially. Why? Because with the abundance of food, their quality of life was greatly increased.
In the customer service realm on the Internet, people data plays the exact same role. Companies using people data to talk to customers as actual humans — and companies who provide their users with a customized, enhanced online experience — will lead the pack.
Obviously treating people like individuals is good — but how do you do it? By aggregating social data over-time.
People data, in general, is not a real-time system, because real-time here doesn’t really matter. What does matter is knowing who your audience likes and follows on social media, what links they click, how long they stay on the site that they clicked over to and how many converted versus bounced.
It also helps to know how different users with different social media preferences act on your site, which ones have bought from you before, if they liked what they bought and if they use or relied on a review system to do so. Then, based on people data combined with on-site analytics, a site can customize experiences for users based how those customers want to use a site.
This reduces confusion, lowers the bounce rate and overall helps marketing teams understand how to treat their loyal customers the same way an offline boutique would: as though they are a member of a shared community with mutual likes, ideas and sentiments. In other words, as a trusted friend.
In all, these three terms spell out the future of the Internet as we will one day know it. The only question left is now that you have all the warmth, mobility and nourishment you need: How will you use it to propel forward?
Higinio “H.O.” Maycotte is founder and CEO of Umbel.
