5 business analytics trends that shaped the start of 2022

Insight to action

In mid-June, longtime analytics vendor Tibco previewed the next iteration of Spotfire, one of its three analytics platforms.

Like many platforms that predate the cloud but are reorienting toward it, Spotfire 12 will include new cloud-native features as Tibco makes its tools more cloud-friendly and eventually completely cloud-native. In addition, it will feature some new machine learning capabilities and capabilities that better enable administrators to govern their organization's data.

But perhaps the highlight of the update will be the debut of Actions, a tool that will enable customers to trigger actions in their operations directly from their analytics dashboards.

Tibco is not alone.

Many other business analytics platforms -- Domo, Microsoft Power BI, Qlik and Yellowfin among them -- similarly enable users to develop integrations between their dashboards and operating systems to enable actions directly from their BI tools.

"The trend that we're seeing right now is this action-oriented story," said Donald Farmer, founder and principal of TreeHive Strategy. "Everyone is going after that, which is good. It's been a long time coming. The ability to take action within your BI tool is a trend of the moment."

Similarly, David Menninger, an analyst at Ventana Research, said that the trend of tying analytics to actions had been a long time coming.

The trend that we're seeing right now is this action-oriented story. Everyone is going after that, which is good. It's been a long time coming. The ability to take action within your BI tool is a trend of the moment.

Traditionally, analytics was read-only. The technology enabled customers to derive insights from their data, but once users had reached a conclusion they were on their own when applying that conclusion to their organization.

Now, that's changing. And turning insights into action is not merely a trend that shaped the early part of 2022 but one that will continue as even more vendors emphasize the capability, according to Menninger.

"We're starting to see vendors offer links between analytics and operational applications," he said. "I expect we will see more of this for two reasons: One, it really helps close the loop on the analytics process, and two, competitive pressures will force more vendors to participate."

More AI and machine learning

While the ability to more easily take action from insight is just now becoming a reality, analytics vendors have been adding augmented intelligence and machine learning tools to their platforms for many years.

ThoughtSpot was an early adopter of the trend, from its inception building its analytics platform around natural language search.

Now many vendors offer natural language query capabilities, and some like Tableau with Data Stories -- a result of Tableau's acquisition of Narrative Science -- and relative newcomer Toucan Toco even are able to provide automated explanations about users' data so data consumers don't have to interpret the data themselves.

Meanwhile, vendors are also adding more easy-to-use machine learning tools such as automated machine learning (autoML) and no-code model development to their platforms so data scientists can more quickly and easily build and deploy data models and business users create and put into action their own models without having to consult a centralized data team.

Qlik is among the vendors that have added autoML capabilities, acquiring autoML vendor Big Squid in October 2021. Alteryx is another.

"Vendors are applying augmented intelligence and machine learning to their products in order to make them easier to use and to help automate various steps in the BI/analytics processes," Menninger said. "They are also bringing to market automated insights, often using natural language processing to explain correlations in the data."

Beyond just enabling queries in natural language and providing automated insights, natural language processing tools are becoming more sophisticated, he continued.

A significant hurdle to NLP has long been the nature of language itself. At first, syntax had to be specific for the analytics tools to understand the user's request and return a meaningful response. Now, NLP features can understand more varied word combinations so that customers can have a meaningful interaction without needing to know code to query and analyze their data.

In addition, many NLP tools only understood one or two languages. Now, vendors are enhancing their NLP capabilities by adding more languages and enabling users to employ more flexible syntax in natural language queries.

"Vendors continue to invest in and expand their natural language processing capabilities," Menninger said.

External data

As vendors add more capabilities by making insight-to-action and AI and machine learning ongoing analytics trends, new trends are appearing among data consumers as well.

One approach that more organizations are choosing is using external data to augment their own data during the decision-making process.

Organizations have troves of their own data, especially those organizations that have been in existence for more than just a decade or two.

Historically, organizations have used that data to find patterns such as seasonal fluctuation and influence business decisions. For example, a historical increase in the purchases of shorts and bathing suits in the spring and summer followed by a subsequent decline in the fall and winter would logically lead a retail outlet to stock more shorts and bathing suits in the spring and summer to meet demand but reduce orders for such items at other parts of the year.

But the more analytics data an organization can use to inform decisions, the better. So while an organization's own data can lead to well-informed insights, adding data that includes partners' data or industry-wide information can lead to even greater insights.

According to Cindi Howson, chief data strategy officer at ThoughtSpot and a former Gartner analyst who hosts The Data Chief podcast, organizations that used external data to augment their data in 2021 outperformed their peers by double digits.

As a result, going forward, external data will no longer be just a luxury.

"Now, in 2022, external data is no longer a nice to have," Howson said. "What we're seeing is that from a business viewpoint, companies that do this get the better leading indicators."

In addition to helping improve business outcomes like increasing sales, organizations can use external data to help create a better customer experience, she continued.

For example, in the instance of healthcare, an insurance provider might have a customer who is diabetic and know when and where that customer visits their doctor. That customer, however, more directly interacts with their doctor, and may be giving their doctor feedback about their medications that the insurance company isn't privy to.

But if the insurance company can access the information given to the doctor by linking to that external data, the potential exists to improve patient outcome and customer experience, according to Howson.

Similarly, in another example, she noted that by linking the data between a consumer product goods companies and retail companies, organizations can achieve better results.

"They're able to create a business ecosystem moat," Howson said, referring to the concept of creating a competitive advantage to protect market share and profits. "The cloud enables this kind of data sharing without moving the data. The worst thing that can happen to any kind of consumer product goods company is that their product is in the wrong place, and this is where sharing data with the retailer and producer form an ecosystem moat."

Cost control

Just as the use of external data is a business analytics trend that gained momentum during the first half of 2022, another trend that is gaining momentum is cloud migration, with many organizations moving their data and analytics from on premises to the cloud.

The benefits of the cloud are significant.

Perhaps chief among them is speed, which amid the uncertain economic climate that began with the onset of the COVID-19 pandemic and has continued through supply-chain disruptions and the onset of war in Ukraine, has become critical.

With economic conditions changing rapidly over the past two-plus years, agile data-driven decision-making has never been more important, and the cloud is what enables the real-time ingestion and processing of data to expedite informed decisions when they're needed.

But the cloud comes with a cost.

In particular, the cost of running live queries on large amounts of data stored in data lakes can be expensive, according to Dan Sommer, senior director and global market intelligence lead at Qlik.

"What organizations are waking up to is that when you only use this technique, they can end up with runaway cloud compute costs," he said.

Meanwhile, the stock market has fallen precipitously this year, inflation has skyrocketed, and many economic observers fear a recession. As a result, organizations' CFOs are getting involved more with analytics and setting up cost governance frameworks, Sommer noted.

Their challenge, just as data governance measures need to effectively strike a balance between risk management and user enablement, is to control costs while still encouraging employees to work with data.

"More cost governance frameworks are evolving as a lot of organizations are seeing spiraling costs," Sommer said. "But you don't want to limit queries. That's where the goodness lies."

One strategy some organizations are employing is examining which queries are most common and effective, so users don't ask as many random questions of their data, he continued.

"Rather than using live query exclusively, you need a data management and analytics approach based on your frequency and latency requirements," Sommer said. "It could be helpful to plot out a 'heat map' of your typical queries, with a mix of in-memory and live-query, and a mix of batch and real time."

Real-time decision-making

At the heart of many analytics trends is an effort to increase the speed with which decisions can be made.

The ability to make fast, accurate decisions has always been important, but the onset of the COVID-19 pandemic made speed an essential part of the business analytics process.

Conditions, whether related to the economy or peoples' health, were changing quickly and constantly. Organizations, therefore, needed their data to be as current as possible and their organizations' decision-making processes to be as nimble as possible.

And that has only continued with the onset of the war in Ukraine, the "Great Resignation" phenomenon and rising inflation. As a result, enabling agile decision-making is a trend that began a few years ago and continued to influence both vendors and data consumers throughout the first half of 2022.

Because of that need for speed, Qlik made the concept of "active intelligence" a guiding philosophy in 2020, and MicroStrategy continues to augment its HyperIntelligence tool to provide insights to users within their workflows. ThoughtSpot, meanwhile, has dramatically boosted the speed of its platform by making it completely cloud-native. 

"Decision-making cycles are shrinking," said Ajeet Singh, executive chairman of ThoughtSpot. "You used to have a few decisions that were made in a top-down manner, and now you can enable your front line to make their own optimized decision."

That agility is critical, he continued.

While it enables organizations to adjust when supply chains are disrupted, as they have been many times since the start of the pandemic, or demand for their services ebbs and flows more rapidly and with less predictability than before 2020, agility also enables organizations to compete with peers that also are investing in real-time decision-making.

For example, in the retail industry, organizations need to make decisions constantly about which items to promote.

There used to be predictable cycles -- that example of shorts and bathing suits being popular on a seasonal basis -- but now demand is less predictable. A consumer might buy shorts or a bathing suit online rather than in a physical location, and because they can make the purchase from the phone or laptop, they might be more inclined to buy the items in November rather than May.

And if the retail company has gathered data on that consumer and knows they might buy something "out of season" if there's a deal, the retail company can personalize their marketing toward that consumer.

"In the retail industry, merchandising used to be done maybe every month or every quarter, but now merchandising happens every hour," Singh said. "Every hour you have to decide whether to promote yellow shirts or red shirts, depending on the need. Decision-making cycles are shrinking, and businesses are becoming very personalized, and that requires a model of micro-decision-making at scale."