From data to insight: the time has come for the enterprise analytics platform

MBA graduates of a “certain age” will relate to the challenge of plowing through page after page of seeming extraneous information in a case study—planted by the teacher to throw you off the scent of the actual issues of the case. The teachers had a reason for this: The business world doesn’t usually present the information you need in an orderly or coherent way.

You don’t know what’s important and what’s garbage. Fast forward to today, and the problem has gotten exponentially worse. Those old printouts seem quaint. This is why companies today have an analytics platform, which often includes an enterprise search solution.

What is the Analytics Platform?

An analytics platform is a comprehensive set of software tools, arranged in a purposeful architecture, that enables an organization to generate insights from its data. It’s a potentially confusing idea. After all, aren’t there are many different data analytics software packages and business intelligence (BI) solutions out there? How are they different from an analytics platform?

The answer has to do with the essential realities of transforming data into worthwhile corporate knowledge – on a repeatable, secure basis. The data industry tends to understate the complexity of the analytics workload. The pitch seems to be “add data, stir, and out come incredible insights!” As most of us know, it’s not that easy.

Getting insights out of data requires having a complete system for data ingestion, management, and analysis. The analytics platform has to handle these functions, but it also needs to operate across the full data lifecycle. Specific architectural approaches to creating an analytics platform include functionality for data access, such as connectors with databases, data visualization, security, and governance.

The platform itself requires a host environment, such as a public cloud platform. However it’s designed and deployed, the key to success will be in the analytics platform’s ability to evolve and continually integrate with new types of systems and data sources. It has to be highly scalable.

The analytics platform should enable data access to all of an organization’s information, regardless of form. It needs to provide data discoverability and analysis of virtually any file type, across all applications. Ideally, the analytics platform will also support natural language processing (NLP), the better to automate interpretations of data and transform it into valuable information. Achieving this goal might mean using machine learning (ML).

Why your data is strategic

We all understand intuitively why data is strategic, or at least we should. Data defines business reality. If we can measure it, we can manage it.

Trends that affect the business emerge from data analysis. We can gauge the true nature of our competitors with effective data analysis. Seemingly minor, or even invisible issues in the marketplace could become pivotal for growth and success—if we have the analytics capability to discover them.

Data is embedded in every aspect of business strategy. Competitive differentiation, for example, a cornerstone of strategy, is all about data.

Data tells us about venues where we can succeed where others have failed. It can reveal what customers want before anyone else figures it out. Data can also provide the basis for strategic advantage over suppliers, another core element of strategy.

The difficulty, however, is that the data itself is seldom able to deliver this level of insight, no matter how rich and plentiful it may be. That takes something extra.

Making data into a real tool of strategy invariably means becoming information-driven. This is partly about the quality of analytics. Truly great analytics turns data into information that helps people make strategic decisions.

An information-driven organization is also one where employees do not have to spend much time searching for the data they need. It’s an organization that reduces what cognitive burden—the extra thought and effort needed to evaluate various data options and arrive at an optimal decision. A well-built analytics platform helps on both fronts. It can perform great analytics, while also putting information into context, so as to surface insights and enable high productivity in the analytics workflow.

Structured vs unstructured data

The ability to work with unstructured data is another “must-have” attribute in an analytics platform.

To understand structured vs. unstructured data, consider the difference between a database table and a Microsoft Word document. The former has data organized neatly in rows and columns. Each field has data in it. That’s structured data.

The database is a highly structured entity. Unfortunately for those who want to use all data to support business strategy, structured data only accounts for about 20% of data in an organization.

In contrast, unstructured data is pretty much every other bit of information out there. It’s the other 80%, which can often be invisible.

The Word document has plenty of data in it, as well as metadata, such as the name of the person who wrote it. There’s no meaningful structure. That doesn’t make the data any less valuable as a strategic asset, however. In some senses, the unstructured data may be even more useful for strategic decision-making.

Unstructured data analytics can reveal extraordinary insights. For example, analyzing social media posts, a form of unstructured data, can show trends in customer sentiment that cannot be found in a database. The same can be said for customer emails or transcripts of customer support calls. These unstructured, sometimes hard-to-get-at bits of data can be extremely helpful in determining what’s really happening inside a business.