There was a time when data analytics was a pipe dream for most organizations. In the '80s, there was very little work done on predictive modeling tools to study the transformation of raw data into meaningful, actionable revenue-generating data analytics. There was no way of accurately forecasting what would happen next based on data that was collected and gathered at the time. The turn of the century brought about an onslaught of data with the emergence of social media firehoses, movement to the cloud, a plethora of media vehicles creating new content, and more.
This brought about a digital transformation in businesses, which began to ask important questions such as – how do you provide digital footprints and provide digital attribution to prove that a media campaign works? Over the last two decades, the focus has moved from merely protecting data to using the data to make predictive insights, deliver actionable intelligence, and generate increased revenue.
"The government support around UAE's Vision 2021 gives people the liberty to try implementing data analytics models and drive toward digital transformation. The hardest part of data analytics, however, is knowing what question to ask. When that happens, analytics becomes front and center. You turn every 'data worker' into a 'discoverer of marginal profitability' … and when that happens, all bets are off on what the opportunities are. It's only going to get more acute. The larger the data sets, the shorter the time frames, more of this activity will start occurring," said Dean Stoecker, Chairman and CEO of Alteryx, in an exclusive interview with AMEinfo.
From use-cases to implementation
Alteryx, which aims to revolutionize business through data science and analytics, works with some of the largest oil and gas firms in the world. From leak detection and price optimization to supply chain management, the use of data analytics has not only changed the workflows of these energy companies but has also helped them grow their revenues.
"One of the used cases that I find most fascinating is the effective use of IoT data coming from centers on drilling rigs out in the ocean. If you have large energy organizations with thousands of rigs around the world with millions of parts, you want to make sure that you can shut the rig down naturally to replace a broken or fatigued part, rather than shutting the rig down unnaturally. It costs a few million dollars to shut the rigs down unnaturally, and a few million dollars more to start it up again after it has been shut down unnaturally. These large energy organizations use machine learning algorithms to predict when to shut a rig down — within milliseconds — when they know a part is about to fail." Dean Stoecker said.
The use of data analytics is not limited to the energy industry alone. A number of large global banks have begun using it to streamline derivates modeling, for next-best-action models for financial advisors, for omnichannel analytics around online and offline banking, for network planning and optimization of ATM placements, as well as equity pricing.
Retailers have begun using data analytics for supply chain, network optimization, risk and compliance, ethics, personalization, and hyper-local merchandising. Dubai Airports is running the airport on analytics with airlines using predictive insights for frequent flyer programs that aid balance sheet improvements. In the automotive industry, it is being used on the factory floors in Detroit, by Ford and Audi, while in healthcare it is being used to solve the world’s opioid crisis.
"The reality is that building an algorithm today is not hard. We see it happening seamlessly. The challenge is that the analysts need to know what model to choose, and how to decision off of an 'R squared' or a 'K-Means Centroid', the distance between one cluster and another cluster, etc. So, understanding the results of an algorithm is something that we are working on – we refer to it as assisted modeling – because we have to continuously amplify human intelligence. Otherwise, we are going to resort to putting everything to a machine and that's not going to end well either. I just don't think that singularity will be met anytime soon," Dean Stoecker said.
The "new oil"
Data analytics is quickly becoming the "new oil" in a hydrocarbon-rich region that is looking to diversify away from its dependence on oil revenues by focusing more on digital transformation and innovation.
"You could wait here for another decade and pump oil all-day-long and it will create economic value, but the way you drive value for the next generation, you have to start today. There will be a lot of countries that will not survive through the digital age. However, with the increased focus on digital transformation, especially with initiatives such as UAE's Vision 2021, the region is primed to become a leader in the age to come," Dean Stoecker said.
A lot of countries and industries are quickly latching onto the data analytics trend that can no longer be ignored or placed on the shelf. In markets such as Norway and Sweden, where wireless penetration is close to 95 percent, 'churn analytics' has become very critical. In the U.S., where healthcare is expensive and the data associated with healthcare is complex and disparate, analytics is going to be the most important driver to bend the cost curve. Gartner estimates that 90 percent of large organizations will have a chief data officer by the end of 2019. This is indicative of the necessity for companies to stay abreast of the times in order to develop and sustain a competitive advantage as well as improve efficiency through optimal use of its data.