The Value of Predictive Analytics and Where We Are Going

Predictive analytics is not new, in fact we’ve been trying to predict the future for thousands of years. The first algorithm can be traced back to the 9^th century, and the ancient Greeks were creating observations to predict weather which ultimately gave us the words “climate” and “meteorology.”  What has changed, however, is how quickly and easily we can create predictions and insights against ever growing amounts of data by making use of computer technology and by enabling algorithms themselves to learn and to develop their own models and predictions without human intervention or direction. This has become commonly known as “machine learning.”

Until now, the science of creating complex mathematical algorithms were squarely in the realm of highly trained statisticians who needed to have deep expertise across three key areas: statistics and mathematical modeling, business expertise, and computer science proficiency. As you might guess these resources were very scare and therefore very expensive—but the value that could be unlocked was so great that many companies invested and benefited from what became known as “Data Scientists.”

But the world of data science is an exclusive one, reserved only for those companies with the significant means required to hire these rare and expensive resources. This is what SAP (KXEN) has been changing for the past 20 years—by simplifying and automating the entire process of data science and making it accessible to people who have the business expertise and a basic understanding of statistics and data preparation.

The Rise of the Citizen Data Scientist

We’re seeing a significant shift in the market today as traditional business analysts and BI Power Users start to dabble into what was the traditional realm of statisticians and data scientists. The prediction is that this group of power users will be the ones to unlock the value of predictive analytics for your average organization that can’t afford to hire large teams of expensive PhDs and professionals in data science, and will allow them to compete against (and sometimes beat) their Goliath competitors.

In fact, Gartner predicts that 40% of data science tasks will be automated by 2020 and that these new professionals (termed Citizen Data Scientists) will surpass the amount of analysis created by traditional data science teams as early as next year.

So Where Is SAP Going, and How Is SAP Supporting this Transition?

First, we’re keeping our commitment to making data science simple by bringing it to the SAP Analytics Cloud so that users will be able to use machine learning technology side by side with business intelligence and business planning capabilities in a virtuous cycle. Imagine a team of business analysts being able to discover the core drivers behind your business, then taking those drivers and building a business plan based on data driven insights instead of guesswork.

Once the plan is in place, users can identify deviations from that plan with easy-to-use visualizations and even see potential deviations based on a predictive forecast—and corrective action can be taken to ensure the plan is met. All of this can be delivered on a single cloud application with total alignment from the boardroom to the shop floor, with up-to-the-second updates available at anytime, anywhere, on any device.

Second, SAP will help your business analysts on their journey towards adopting machine learning in order to drive business results. There will always be a continuum of users who are on a journey towards adopting new technologies and techniques. Here is how we will help each of those users:

• Your average reporting consumer doesn’t want to build predictive models and would rather not even have to look at a report. The average user wants to be told where the problem is, what can be done to fix it, and how to get started. By embedding the results of predictive models designed by other experts directly in the applications where knowledge workers are, we’ll give them the prescriptive insights they are looking for.
• The traditional BI report designer, or business analyst, who is looking to understand how machine learning can be used but doesn’t have any training will benefit from the Smart Insights and Smart Discovery capabilities on SAP Analytics Cloud. Without having to build a model, or even know any statistics, we will give this user detailed insights about their data and the patterns that are hidden within that data allowing them to identify key areas for potential insight and deliver these directly to their end users.
• As business analysts become more educated and comfortable with key data science principles and start to think of business problems in terms of predictive problems, we’ll provide a guided experience that allows them to build reusable models against trusted data sets that can be consumed by all users.
• The most advanced users who understand more detailed statistical concepts like data manipulation will benefit from the automation of data acquisition and feature engineering and can create trusted and reusable data sets for their organization. They will also be able to help validate the work being done by less experienced users and help to scale the application of machine learning across an organization.

By supporting every user along their data science journey, SAP is empowering the consumer, the business analyst, the citizen data scientist in training, and the mature citizen data scientist to bring the benefits of data driven insights across an organization.

Learn More

• Learn more about the predictive analytics capabilities in SAP Analytics Cloud by reading our other Machine Learning Thursday blogs.
• See and experience for yourself what’s next for predictive analytics at SAPPHIRE NOW June 5-7, 2018.