From environmental disasters to financial crashes and political shocks, we live in a world that is increasingly difficult to predict. It’s a fact that is compounded by the accelerating pace of change of the digital age. And yet, against this backdrop, businesses need to be able to make decisions and make them quickly.
Making the right choice requires a company to understand every aspect of its business—in the past, the present, and the future—and to recognize the value of the data available to them and what it tells them about their business. Ultimately, the aim of analytics within the enterprise should therefore not simply be to report on what has been, but to enable everyone at every level of an organization to make decisions with confidence.
It’s a big ask of any company, but as we head deeper into 2019, what are the analytics tools, features, and functionalities we can expect to see to help businesses do exactly that?
While the adoption of predictive analytics methodologies is certainly increasing, this change has been mainly driven by their IT specialists, with business users having to make requests for (and wait on) their reports. With demand for data scientists outstripping supply, however, many companies are looking to bridge the gap by introducing self-service capabilities to their employees. It’s a trend that benefits both parties – while the business users can access capabilities previously out of bounds, by not having to spend their time on such tasks, the data scientists can focus on more complex and higher value projects.
As such, what was once known as “advanced analytics” will rapidly become part of the standard toolset of everyone from marketing professionals to accountants. According to research company Gartner, this shift will mean that “by 2020, more than 40 percent of data science tasks will be automated, resulting in increased productivity and broader use by citizen data scientists.”
Business Intelligence (BI) will evolve to include advanced analytics capabilities such as automatic data discovery. And while such developments certainly add to the tools available to users, considerably improving their ability to make strategic decisions, they also go one step further: They prevent users from falling into the bias trap, whereby data discovery justifies an outcome rather than reveals a new insight.
Left to our own devices, it is an inherently human trait to find only what we are looking for when analyzing data. Consciously and unconsciously, we guide the process and sort the data to get to the information that confirms what we expected. But by being so focused on what we think should be there, we can also miss important trends.
Smart analytics tools avoid this by actively drawing the user’s attention to information that might prove important but that could otherwise go unnoticed. Behind the scenes, a set of machine learning models provide an overview of significant patterns, outliers, and key influencers of the business that help users really understand what is going on in their business. By transitioning from a passive system (“here’s some data, interpret it how you will”) to an active system (“have you seen this unusual development over here? It looks like it is being caused by this…”) analytics practices are actively helping users understand what is happening now, why it is happening, and how that will impact future results, all ultimately improving the speed of decision making.
The Consumer-Grade Analytics Experience
Similarly, thanks to advances in areas such as natural language processing, solving business problems should be as easy as “Googling” any other question. Just as no one needs to understand the programming behind a search engine to be able to use it, no one should have to first learn coding to get the answer to the question they are looking for in their analytics solution. We can expect to see tools that enable users to do just that—benefiting from conversational technologies to get the answers to questions such as “what are the top ten stores by sales revenue in Germany?” by simply typing the question.
Finally, tools such as automated model builders represent another important development. By giving business users access to capabilities that allow them to solve standard predictive modeling tasks, they can leverage the tools of a data scientist without having to actually become an expert themselves. Incrementally introducing and exposing users to such concepts—and all without any significant upfront training requirements—also plays a key role in helping to establish a data- and machine learning-driven culture in an organization.
In 2019, we can expect to see augmented analytics methodologies being used pervasively across companies. From the boardroom to the shop floor, analytics has become a tool that can be accessed by everyone. As users and as people, we bring our own unique perspective to our analysis of the data. It is exactly this combination of such powerful artificial intelligence and the inherent creativity of the people who use it that ultimately enables us to make decisions faster and with greater confidence than ever before.
Gerrit Kazmaier is senior vice president of SAP Analytics, Database, and Data Management at SAP.