Sergey Nivens - stock.adobe.com
The COVID-19 pandemic heightened the importance of data-driven decision-making at all levels of organizations in 2020, and the analytics trends expected to drive the market in 2021 will continue to extend the reach of business intelligence.
Before the pandemic, many organizations viewed analytics as a luxury. But once COVID-19 began to spread in March 2020 and led to stay-at-home orders that essentially shut down local economies, the need to find efficiencies, manage supply chains and even find new avenues of business through analytics became critical.
Organizations already undergoing digital transformation accelerated the process, and many of those that hadn't yet begun to adopt data-driven cultures finally understood the need.
They made ease of use a priority so more business users -- not only data scientists and data analysts -- could work with data and took steps to increase the data literacy of end users.
"I think 2020 made so many people realize the digital transformation journey is not going to slow down," said Sudheesh Nair, CEO of analytics vendor ThoughtSpot. "It's not going to come back to where it was [before the pandemic], and that's good."
So the analytics trends that analysts and executives think will be of utmost importance in 2021 are those that build on the digital transformation that began over the past few years but accelerated in 2020 because of the pandemic.
They're those that address efficiency to speed the time it takes to complete tasks throughout the analytics pipeline, and through ease of use enable more users throughout the organization.
COVID-19 exposed the need for real-time data.
Healthcare organizations needed to know how and where the virus was spreading in order to prepare for potential case surges, local governments needed to know the risks in their municipalities in order to make decisions about their economies, and enterprises needed to know whether their customers were still coming to them or not.
The quarterly and monthly reports many relied on previously suddenly became irrelevant.
Enterprise data collected in February or March had no meaning in April or May when suddenly few were shopping for anything other than food in person, and data collected over the summer when COVID-19 cases were low and warm weather opened possibilities for certain businesses had no meaning in the fall when the spread of the virus again spiked and the weather turned cold.
Organizations need to know what happened five minutes ago, not five months ago. And year-over-year data is utterly irrelevant when what happened 12 months ago was before the pandemic.
Continuous intelligence became vital in 2020, and will remain among the significant analytics trends in 2021.
"Since the pandemic arrived, we've seen a surge in the need for real-time and up-to-date data," said Dan Sommer, senior director and global market intelligence lead at Qlik. "What is usually fairly stale -- quarterly business forecasts, for example -- is fleeting and mutable now."
Similarly, Kimberly Nevala, AI strategic advisor at SAS, said that the pandemic exposed the weakness in relying on historical data and seemingly predictable patterns.
As a result, organizations will rely more on real-time data.
"In 2021, organizations will bolster investments in traditional analytics teams and techniques better suited to rapid data discovery and hypothesizing," she said.
Meanwhile, putting continuous intelligence to good use relies on being alerted to changes in data.
It's those push notifications that have the potential to enable healthcare organizations to know a spike is starting and they need to prepare. Alerts can also notify organizations of potential supply chain problems, and enterprises about changes in customer behavior.
Based on those alerts, organizations can see any changes quickly and react much earlier than they could if they waited for the next monthly or quarterly report.
"The infrastructure and applications are available, enabling a gradual transition to active intelligence," Sommer said. "That will be a big factor in helping enterprises preact."
Continuous intelligence has the potential to make organizations more efficient, able to prevent losses and drive revenue increases by reacting to changes more quickly than before and taking proactive action based on fresher intelligence than in the past. Process automation is an analytics trend that has similar potential to make organizations more efficient.
Once the mundane task of IT departments, data management tools are now able to automatically do in a short amount of time what it once took data scientists weeks and months to do. Tools are able to automatically load data into data lakes and warehouses, and once there do all the transformation and organization needed to prepare the data for exploration.
As a result of process automation, data is ready for analysis much more quickly than when humans do all the data preparation work. Meanwhile, those humans whose time was dominated by the painstaking data management process are freed up to develop models and do other work that takes advantage of the prepared data.
Sudheesh NairCEO, ThoughtSpot
"Reacting fast has become critical, and business processes are at the center," Sommer said. "Business process management has been around for decades. What's new is that we can not only model it but also mine, automate and optimize a process via technologies like robotic process automation."
Beyond preparing data, tools now automatically monitor key performance indicators for changes and any other problems that may arise, saving organizations from potentially reacting too late to evolving situations.
"New data observability tools monitor, detect, predict and resolve all the issues that cascade from source to consumption," said Kevin Petrie, vice president of research at Eckerson Group. "These observability solutions help make data pipelines faster and more reliable. Site reliability engineers, platform engineers, data engineers and architects -- not to mention the business owners they serve -- all stand to benefit."
Automation even has the potential to make consuming data easier.
While data scientists and analysts have the data literacy skills to interpret data and make data-driven decisions, most employees don't have the training to authoritatively interpret data and make critical decisions.
Data storytelling tools are able to automate the explanation of data, creating a narrative about the data in natural language. Meanwhile, Glen Rabie, CEO and co-founder of Yellowfin, predicts that BI vendors will develop tools that take a guided approach and shepherd the end user through the process of working with their data.
"New simplified user interfaces will allow business users to interact with data in a more guided approach, allowing them to reduce insight time with minimal analytical skills," he said. "Automated analytics will shift from the enterprise domain toward software vendors who will embed these capabilities and enable mass adoption through their customer base."
The evolution of AI
Augmented intelligence, a subset of AI, is the trend that took business intelligence beyond data visualizations into what many regard as the third generation of analytics.
Some platforms are now able to take on the data analysis itself. Meanwhile, the need to painstakingly write code to develop applications has been eased by low-code and no-code application development tools and advances in natural language processing (NLP) have enabled business users without backgrounds in data science to ask questions of data and engage with data.
AI features will continue to progress in 2021, and in particular NLP has the potential to extend the reach of analytics to new users.
NLP has existed in some form for years, but it's been limited by the complexities of language itself. Worldwide, more than 5,000 languages exist, but even within those most widely spoken are words that sound the same but have different meanings, words spelled the same that have different meanings, and words that mean the same thing but bear no visual or auditory relation to each other.
But NLP technology is advancing, and becoming effective.
"Natural language has started to revolutionize the way people learn from and interact with data," said Mike Leone, senior analyst at Enterprise Strategy Group.
Data storytelling platforms, for example, are now able to query data and develop narratives to explain the results in plain language.
NLP is not, however, widely adopted yet, according to Leone.
"Natural language integration and usage is at the early stages of adoption when it comes to business intelligence," he said. "As organizations continue to look for ways to enable all end users to better leverage data, expect natural language adoption to explode in the coming year."
But beyond the technology itself, AI is expected to mature in other ways in 2021, and Petrie predicts organizations will begin to monetize their AI assets in 2021.
Enterprises need AI models to optimize their business, but not all have the wherewithal to develop their own models. The number of those that do have the capacity to develop models, meanwhile, are increasing, and they're looking for ways to generate profits from their capabilities.
"These forces of supply and demand can find their equilibrium in AI marketplaces, which help enterprises and individuals exchange AI models for profit," Petrie said. "Business managers can find and buy models, data scientists can create and sell them, and developers can use and integrate them with applications."
Nevala, meanwhile, predicts new government regulations expected to be developed in coming years will spur more AI adoption.
The relative absence of regulation, she said, has kept organizations from developing new features out of fear they'd violate yet-to-be-written laws and overstep privacy bounds.
"Emerging regulations will support AI adoption by providing explicit guardrails for normalizing and contextualizing risk," Nevala said. "This will enable organizations to more confidently navigate the uncertainties inherent in these probabilistic learning systems."
Nevertheless, she added, they need to be mindful that regulations may vary across different regions.
The standbys: Data literacy and embedded BI
Data literacy and embedded analytics have been rising trends for a few years now, and they're expected to continue to grow in importance and adoption in 2021.
Embedded BI enables self-service users to work with data anywhere, and at any time by opening analytics up beyond just the traditional data analysis stage of the analytics process when an analyst looks at a chart or dashboard and makes a decision based on what they're seeing.
It can deliver information to users as they scroll over certain words or numbers on their screens. It can alert users as they prepare data. It can even enable frontline employees at the point of sale as they engage directly with customers.
But according to Leone, despite its potential, embedded BI remains more an idea than a reality.
"Over the last several years, vendors have emphasized the importance of self-service BI to help enable the democratization of analytics, but fact of the matter is that it hasn't picked up as much steam as vendor marketing has portrayed," he said. "Embedded analytics will prove to be the ideal approach going forward to truly enable democratized access to analytics and business intelligence."
Similarly, Rabie said that BI in context -- not just when running queries and looking at charts and dashboards -- will be a trend in 2021.
"Expect to see businesses embrace the power of contextual analytics, allowing them to more easily dive into analytics at the point of decision-making while integrating operational workflows with guided analytics," he said.
Data literacy, meanwhile, is another current trend that will continue to gain momentum in 2021.
While there are storytelling tools that automate data interpretation and enable more end users to work with data, they have limits. They're meant to inform decision-making, not actually make the decisions themselves.
Data literacy, therefore, remains integral to the decision-making process. And never more so than now when data is changing constantly and many organizations are struggling to survive amid the pandemic.
"Increasing the data fluency across an entire organization is critical, and the answer is never going to be to send out more reports or assign a BI person to every function," Nair said.
Data is crucial to the healthcare organizations now dealing with patient surges as COVID-19 cases increase again after subsiding during the summer. It's crucial to the government agencies making decisions about whether to close down parts of the economies again. And it's crucial to enterprises as they make decisions about how to stay afloat.
"This is going to be a year in the world of data analytics where I predict that increasing the data fluency across the entire organization will be a necessity," Nair said. "At least that journey will start in 2021."