Analytics used to be a niche specialty for enterprises with deep pockets used.
No more. Vendors are working hard to democratize analytics. Putting tools in the hands of all users -- not just specialists -- is also changing the way businesses operate.
Democratization of Analytics
“We are at the stage of democratization or humanization of business. Under this movement, anyone in any business can utilize analytics for a good purpose,” said Nelson Petracek, global chief technology officer, TIBCO Software Inc.
But will this oversimplify the role of analytics? “No. As long as we provide insightful interfaces for wider groups of users, then the use of data is always beneficial,” Petracek said.
He believed that TIBCO Software’s primary role lay in providing these interfaces.
“Data can be utilized by everyone in a company to increase productivity and improve decision-making. Better use of data ultimately has a positive impact on the bottom line of any business,” Petracek said.
Still, companies faced several challenges in becoming analytics-driven.
One major challenge is the science behind asking the right questions. It may be trivial for analytics specialists. “However, most enterprises are rather untrained in making use of the data; this is not due to any wrongdoing on their part but stems directly from the difficulties in working with data, data that is more often than not hard to drag out of the silos where it is kept and sometimes hidden,” he said.
Another challenge lies in data preparation. In today's data-driven environment, the risk of getting the wrong insights from biased or noisy data is genuine.
“The difficulty of collecting and preparing data for use with analytics is often underestimated. There is a lot to think about - you need to look at the sources of the data, how ‘clean’ those sources are, how fast the information is generated, what transformation or pre-processing or enrichment is necessary to make the data useful, the time span of the data’s relevance, and the context in which it was generated,” Petracek said.
But the biggest challenge is finding the right people -- especially in the Asia Pacific region.
“The key, common hurdle that companies face is the absence of necessary skills to lead and execute the analytics capability,” Petracek said.
This gap in talent will only get worse as companies compete using machine learning.
TIBCO Software sees these challenges as opportunities. The company is inking partnerships with local universities to increase the talent pool.
The company is also creating tools for the new “citizen data scientists” category. It includes part-time analysts to business users who are analyzing data.
TIBCO Software believes that by putting the right tools in the hands of these users, analytics use will soar. And so far, the signs are good.
"AI/ML is becoming more embedded in ‘everyday' activities, and there are many examples of scenarios where it is often difficult to distinguish certain actions as automated, AI-driven, or human-led," Petracek said.
Rethinking the Proposition
Internally, TIBCO Software is rebuilding its value proposition and business model. The company is already making its tools usable by various personas and skill levels.
“Not everyone needs to be a data scientist. We are also focused on the complete analytics lifecycle, from data collection, data prep, data analysis, data visualization, straight through to model operationalization, which presents challenges and capabilities that are often larger than just the models themselves,” he said.
TIBCO Software is also strengthening its industry collaborations. The recent announcements with Amazon Web Services and Microsoft highlight this. These aim to “improve how analytics are utilized to solve today's business challenges and to support digital innovation,” Petracek said.
Lastly, TIBCO Software is investing in community development and embracing open source. The goal is to share “best practices, developed components, extensions, and other technical or vertical-specific capabilities across the globe.”
“TIBCO Software's goal is not just to provide one or two pieces of the analytics puzzle, but to focus on providing organizations with a set of capabilities that cover the complete analytics lifecycle,” Petracek said.