Data Scientists Explain Why You Don’t Need Data Scientists [Video]

We don’t like to throw around the term “smart data.” It seems to so many like a lofty idea, perhaps a dream product for digital media companies and their audiences alike. And we agree — it is all those things.

So, we nabbed our in-house data scientists, convinced them to put their genius thoughts on camera, and got them to explain to you exactly what it is that makes Umbel so crazy smart that both us and our clients call it a smart data platform as though that’s just the industry norm. Because, well, we think it should be.

Umbel’s Data Scientists Talk Big Data from Umbel on Vimeo.

To give a bit of background, smart data is an extension of big data and big data became a big deal when suddenly so many digital data points were being collected (think via search queries, credit card purchases, social media likes and posts, etc) that no single computer could process them.

To be clear: there has always been “large data,” meaning data sets so big that no single human, or group of humans, could feasibly mine the information for relevant nuggets in a reasonable amount of time. But a computer could — and then spit everything out into a spreadsheet that a data scientist or miner could more easily navigate, connecting the dots across previously hidden relevancies.

“Umbel takes care of the details of managing a huge amount of data.”

Today, one computer isn’t enough. To solve this problem, tech and dev teams have invented cluster batching, which is the use of multiple computers working in sync, or as a team, to process very large amounts of data in a timely fashion. Cluster batching, as you might imagine, is relatively expensive, given that you need multiple, perhaps an infinite, number of well-functioning computers.

Hadoop, though, has solved much of this problem — and put cluster batching into the cloud. It’s an impressive feat, to be sure. But, like the data processing of old where you still needed a miner to pull out relevant nuggets of information, Hadoop requires the same type of Ph.D. hire — or a platform like Cloudera or Hortonworks to sit atop it and better visualize the findings for executives.

We Stand By Our Claim: You’ll Never Need a Data Scientist

As one of our dev guys said, “Umbel takes care of the details of managing a huge amount of data.”

Here’s what’s going on behind the scenes: Umbel attacks big data from the audience intelligence angle, sorting through and intelligently filtering the online identities of your users. Where do they click on your site? When do they convert? What other brands do they like on social media? Which brands might you be able to partner with in ways that make sense for you, the brand and your customer?

Umbel can answer all these questions, and more. And it does so without ever needing you to hire a data scientist.

Instead, the platform visualizes your data points according to what business outcomes and audience metrics are important to you. Then, the platform offers one-click ways to start positively affecting your bottom line, in real time.

For us, it’s about visualization, curation and organization. So that when you decide to act on what you see — you get the ROI of your dreams, not just the platform.