kantver - Fotolia

Future of analytics will close gap between data, decisions

As data increases in complexity and volume, a gap is growing between data intake and data-driven decision-making that analytics tools will need to address and aim to close.

A chasm remains between collecting data and using it for decision-making, so the future of analytics will be about bridging that gap, according to Peter Bailis, founder and CEO of Sisu Data Inc.

Speaking during a recent webinar, Bailis explained that organizations have the data they need to make informed decisions about their business, but what they lack are the tools to fully use data to make the decisions that will have the most impact both in the moment and for the future.

Older business intelligence tools are equipped to demonstrate what is happening in a given moment, added Joel McKelvey, Sisu's vice president of product marketing, but they are not able to explain why that something is happening or what to do about it.

Meanwhile, organizations want more analytics and also more from their analytics.

The problem

According to Gartner, overall analytics adoption is expected to grow 15% by 2023. Yet, according to McKelvey, a study by McKinsey & Co. said only about a third of decisions were deemed both fast and of high quality.

The future of analytics, therefore, will be about making more decisions based on data and ensuring those decisions are based on good data and wind up with good results.

That means developing tools that augment human capabilities and enabling users to get beyond just a view into what is happening in the present to understand why something is happening, and then what they should do about it, according to Bailis.

"Organizations, increasingly, have the data to make decisions, but it's very challenging, given today's toolkit, to actually do so," Bailis said. "It's what we call the decision gap. Organizations have the data to inform decisions but lack the time, energy and resources to put it to work."

Screenshot of Sisu Data executives speaking during a webinar.
Discussing the future of analytics are Sisu Data's Joel McKelvey (left), vice president of product marketing, and Peter Bailis, founder and CEO.

The data that organizations are now able to collect is far more complex than the data they collected in the past, according to Bailis. In addition, there's far more of it, with the amount of worldwide data expected to grow from 44 zettabytes at the start of 2020 to 175 ZB by 2025, according to IDC and Seagate Technology.

Without tools that augment what humans are capable of doing, getting real value out of complex and exponentially growing amounts of data is time-consuming. Depending on the analytics project, it can take days, weeks and even months for humans to build, train and deploy models on their own.

That time lag, meanwhile, can render the projects meaningless, given how quickly conditions can change.

"The legacy approach to using data really starts to break down at scale with complex data," Bailis said. "Having humans in the driver's seat of the process is really slow. And when the answers come back, they're often incomplete, because with complex data, you can only look at a limited number of factors at any point in time and you're going to miss trends and factors in the data."

The result is that the data winds up not being actionable, he added.

"The gap between data and decisions, driven by this legacy approach, is the state of where we are today," Bailis said.

Organizations have the data to inform decisions but lack the time, energy and resources to put it to work.
Peter BailisFounder and CEO, Sisu Data

Similarly, McKelvey, who noted that Gartner reported only half of all decisions are driven by analytics, said older analytics capabilities are no match for the current volume and complexity of data.

"Decisions are not fast, and they are not as good as they could be," McKelvey said. "That has a real impact on business at every level. The tools that we have been using were sufficient for the volumes and complexities of data we used to have."

The solution

Enterprises need new tools.

More automation will be part of the future of analytics and is one key to closing the decision gap, eliminating human involvement in certain tasks so they're freed up to concentrate elsewhere. So, too, will be tools that augment human capabilities.

Some analytics vendors are already adding capabilities that automate certain tasks, particularly associated with data management. For example, Tableau Software offers Prep Builder and Microsoft's Power BI includes Dataflows.

Meanwhile, augmented analytics capabilities such as alerting, guided analysis and natural language processing are becoming more commonplace, with Tellius among those offering guided analysis capabilities and a host of vendors including alerting and NLP capabilities.

Contextualization -- the ability to automatically give context to data by showing why something is happening and what should be done about it -- is another technology that will fuel the future of analytics.

Sisu, a startup founded in 2018 and based in San Francisco, is among the vendors focused on contextualization with a platform that monitors data sets for changes and then automatically alerts users and explains those changes. Other vendors like Yellowfin and Narrative Science, meanwhile, are attacking contextualization through data storytelling.

The future of analytics will be providing "the right content across all of this incredibly live data so that you get the relevant insights surfaced at the right time, and it's as automatic as possible so that data teams can focus on key factors and not be caught up in the static or not relevant data," McKelvey said.

Similarly, Bailis added that the capabilities that will close the decision gap will be those that advance the capabilities of self-service BI.

The future of analytics will be tools that enable end users to surface insights without the assistance of analytics teams, ask follow-up questions of the data without needing help, do so in a single environment that doesn't require toggling between systems and screens, and automate mundane tasks to free up data teams for human endeavors.

"Invest in enabling your team to self-serve," Bailis said. "Look for tools with a low threshold for getting started, but a high ceiling for experts. A second principle is … we can accelerate analysis for the routine parts and let people spend time on the parts that are uniquely human.

"By accelerating data exploration and data analysis, we let people do what they're best at and let machines do the rest," he continued.

Next Steps

HomeToGo turns data into action with Sisu analytics platform

Dig Deeper on Business intelligence technology

Data Management
Content Management