Because companies crave a competitive advantage and desire better insight into their data, more and more are adopting analytics and machine learning in the marketplace. Gartner predicts the business value created by artificial intelligence (AI) and machine learning will reach $3.9 trillion in 2022.
To put this into perspective, we invited Jeffrey Silverman, senior manager for Grant Thornton, to join us on the Oracle Analytics Advantage Podcast. As a leader in its Business Analytics practice, Silverman advises clients on the advantages of implementing cloud-based analytics embedded with machine learning to gain a competitive advantage.
"The industry is way beyond early adoption," Silverman notes. "There are so many companies out there that see the value in cloud-based analytics that we are in the early majority. And when you are in that position, you're going to reap the benefits and be ahead of your competition."
Silverman is also a Lieutenant Colonel in the military reserves and is an intelligence director for the US Military where he leads the Big Data Initiative for US Strategic Command. He quips that there is such a compelling argument for using cloud-based analytics to derive insight from data that even the US government has adopted the strategy.
"If we're already talking about cloud analytics, then a lot of people should be," Silverman jokes.
Silverman divides the analytics world into four sections. There is descriptive (What is out there? What's going on?); diagnostic (Why did this happen?); predictive (What will happen next?); and prescriptive (How do you make something happen?). He considers prescriptive the most sought after, yet most difficult to produce because it requires the most data and the most analysis.
"If you want to get beyond dashboards and reports and into predictive and prescriptive analytics, you absolutely need to adopt cloud-based and embedded machine learning analytics," says Silverman. "But I will caution if you don't have your data in order and you don't have your descriptive or diagnostic analytics in order, and you try to feed this beast of machine learning…garbage in is garbage out."
Getting to that blend of descriptive, diagnostic, predictive, and prescriptive analytics has great potential. An example of this is the practice of dynamic pricing.
"Hotel websites and other portals like VRBO need to adapt their pricing schedule to accommodate for demand," says Silverman. "Low demand means the price is low. High demand means the price gets higher. Airlines do this too. Machine learning adapts these business rules and applies them to a new hotel or a new location or a new flight route—much faster than if you waited for an individual to do this manually."
Taking a diagnostic approach can also reduce risk. If there is an emerging threat, for example, having cloud-based tools to mine unstructured content and unstructured data allows companies to react faster to customer concerns.
"If there is a choking hazard in a toy, wouldn't it be great to troll through the various reviews that are on a product to see if maybe a customer posted it on their Amazon rating," Silverman says. "If data discovery tools allow you to do that and go through something that is nonstandard data, it's going to allow you to rapidly address an issue before it becomes truly destructive to your organization."
The next phase is artificial intelligence and a layer of machine learning running over AI. You want to make decisions quickly and refine the roles that these layers play in a business analytics environment, allowing them to learn from their past mistakes. Whenever you want to be fast and potentially prescriptive, embedded machine learning will give you an advantage.
"One of the things we are looking at is how profitable is a particular product by historical trend," Silverman says. "Then we take that trend and overlay it over a calendar and ask where we think we are going to be in the future based on consumption."
Check out the entire conversation on the Oracle Analytics Podcast (click the icon below to listen):
And if you ever needed more convincing about the value of cloud-based analytics, the video below is our interview with Joseph Coniker, a principal at Grant Thornton, who explains how Oracle Analytics Cloud enabled Grant Thornton to integrate enterprise performance management and financial ERP data together to offer robust reporting analytics solutions for its customers.