Every so often there are key words and phrases that get big play around the media, and before you know it, they are in use almost everywhere. These terms include irrational exuberance, outside the box, deliverables, skill set, metric, multitasking, bandwidth, selfie, spyware and more than a few others.
I’m sure during the last few years, you’ve heard those words and phrases time and time again… so much so that the meaning behind those words became dulled, if not lost.
There’s another term that has been gaining in popularity during the last year or so, and believe you me, it is one that will impact your very life — big data.
I know you’ve heard those two words — Big Data — before, and it’s been used by companies like Cisco Systems, Facebook, Google, Oracle, IBM and others. In this age where more and more things are connected, monitored and captured, what is big data? How is it used? What does it mean for business and for the consumer? How will it change things?
These are the questions we need to understand, and joining me today on PowerTalk to sort through all of it is Arnab Gupta, Founder and CEO of Opera Solutions. As Arnab points out, it’s not just you and me that are just figuring out what big data is, companies are in the same boat — they know they need to spend on it — and they are, but how do they get value from it? How do companies use it to make better decisions, work more productively and take better actions? How does a company go from spending on big data to making it a competitive advantage for their business?
Arnab and I not only discuss all of this, but he shares how he thinks big data will force companies like Oracle, IBM, SAP and others to expand and change their capabilities, lest they risk becoming the next Eastman Kodak (KODK) or the next Blackberry (BBRY).
Arnab, what is the pain point that Opera Solutions addresses?
Organizations are now starting to understand the potential that Big Data has to drive unprecedented improvements in their business. But they are struggling with how to turn raw Big Data into real value in their organization. Companies are asking themselves, “how can we reliably and efficiently find patterns in all this data that can help us predict what customers, markets, machines, and systems will do in the future? And once we find these patterns and make the predictions, how can we integrate this into the workflow, so that people on the frontlines can make better decisions, work more productively, and take better actions?” This is the pain point that Opera Solutions addresses – reliably and efficiently extracting predictive patterns from Big Data, and turning them into suggested actions, next product recommendations, scores, alerts, etc. that help front line employees do their jobs better.
What is big data and why do businesses and how fast is it growing? Do you see that pace of growth accelerating given what Cisco CEO John Chambers calls “the Internet of Everything”?
Simply put, Big Data is the explosion of digitized data – created by people, machines, sensors, etc. – that acts as an audit trail, capturing what is happening in and amongst humans, machines, markets, natural systems, and the like. The pace of growth will continue to accelerate as:
It becomes cheaper and easier to fit machines, the natural world, and people with sensors, equipment, gadgets, and the like that capture more and more of their movements and changes;
The science of machine learning continues to advance, making it easier to extract highly accurate anomalies and predictive patterns from this flow of data;
Organizations become more adept at using the output of machine learning algorithms to power “use case models” that truly drive greater efficiencies and effectiveness in their operations.
What does this mean for companies — General Electric, Starbucks, Expedia, American Express — and how will it affect their business?
Companies that understand the potential of Big Data and are on the leading edge of extracting and using its predictive value will be huge winners. Just as Wall Street investors value and reward “brand equity,” they will also begin to assess and value “data equity” – that is, what’s created when a company has access to unique data and is uniquely qualified to leverage this data.
In terms of how this will affect specific companies’ business, the impact will be ubiquitous and, in a funny way, so integrated into the “way things are” that it will be invisible. We no longer ask “how does the Internet impact businesses?” — the answer is, it impacts it on every level and therefore is just part of the status quo and the fabric of business life. This will soon be the case with Big Data, machine learning science, and the flood of predictive intelligence – it will just be part of life. The term Big Data is probably going to disappear, in fact.
How does this fit with Opera Solutions’ tag line — profit from big data flow?
The central truth of Big Data is not that it’s big, but that it contains (hidden amidst much dreck) more predictive information than ever before available since the dawn of mankind. We now have the cheap and highly effective computational power, plus the scientific techniques, to begin to extract and use this predictive information.
The next wave of Big Data will be all about extracting and using its value to drive profit, productivity, and competitive advantage. That’s why we chose this tagline. In addition, we chose the word “flow” because we want the market to recognize that today’s machine learning predictive models should not be fed by static data reserves; they need a constant flow of data in order to perform optimally – because they’re looking for patterns that emerge over time and across different data sources.
With such an explosion, is the competitive landscape increasing? How does Opera Solutions differentiate itself?
Competition is no doubt increasing in this area. We have not come across companies that directly compete with us yet, but we expect that we will. There are companies emerging that seek to automate parts of the data scientists’ tasks, particularly in the initial data wrangling, organization, and variable identification part of things – that’s a part of what we do.
There are companies that try to automate some of the machine learning modeling by building in advanced algorithms – that’s also a part of what we do. There are companies that try to take analytic output and visualize it better, so people can use it to make better decisions – that’s also a part of what we do.
Our differentiation is that we do all of these things together – and no other company that we have seen is taking the horizontal approach that we’re taking. Our Signal Hub™ product is an example of this horizontal approach: it ingests many different data sources and organizes them; uses a range of modeling techniques to extract the predictive variables (or “Signals”) from this data; stores, manages, and updates these Signals as new data flows in; and makes them easily available for scientists to use in predictive models for specific use cases.
Do you see a shift in how companies like Oracle (ORCL), IBM (IBM), SAP (SAP), and others will have to conduct their business? How is Opera Solutions preparing itself for that change?
Companies that have built their business on a notion that data is something that can be stored and warehoused – that data is a relatively static, structured, and inside-the-enterprise commodity – will need to rethink this. I don’t think that data warehouses are going away any time soon, but the market is shifting toward much lower-cost alternatives. CIOs are asking, “could we use Hadoop instead of investing multiple millions in a new warehouse?”
CMOs and other functional managers are asking, “how can we incorporate extra-enterprise sources of data like social media, or incorporate data from inside the organization that isn’t captured in a warehouse?”
So, I expect that Oracle and SAP will absolutely have to rethink and expand their capabilities. They have the financial resources and the installed customer base to win in a new setting – but they have to do something very hard, which is to move from an old paradigm in which they were hugely successful, and embrace a new paradigm with unknown and difficult new challenges and great uncertainties.
I believe they’ll figure out how to do this – because the lessons of companies, such as Kodak, that failed to do so are right in front of them. As for IBM, I think they’ve taken many steps to move into a Big Data world, not the least of which is their latest move with Watson. They have the bankroll, the strategic vision, and the marketing capabilities to do well – however, they are also encumbered by their sheer size, as well as by their existing infrastructures, processes, and legacy businesses. The move to spin off Watson as a separate entity was, in my opinion, a smart one. We are looking at it with great interest.