According to Gartner, more than a third of large organizations will have analysts practicing decision intelligence by 2023. This is the next step for organizations looking to establish a cohesive decision strategy across people and departments.
Decision intelligence software uses analytics and machine learning to augment decision models. It's often viewed as a marriage between data science, decision theory and social science.
What is decision intelligence?
In plain English, decision intelligence software uses analytics to help employees, customers or business partners make decisions by offering them data, analysis and predictions when and where they need it.
As decision intelligence becomes a core part of business processes, decisions get made faster, more easily and less expensively than before. The idea isn't to replace people, but to help them make better and more consistent decisions, said Anand Rao, partner and global AI leader at PwC.
Often when individuals make decisions, a bad day -- or some other unrelated personal factor -- can easily skew the decision-making process. And people have different backgrounds and different perspectives they bring to their jobs that may influence their decisions.
"Companies use training and other mechanisms to try to get some consistency," Rao said.
To reach more consistency across teams, organizations can use decision intelligence to provide people with guidance, recommendation and evidence to make decisions.
"It gives you an audit trail and consistency," Rao said.
Analytics and decision intelligence
To add decision intelligence capabilities to their analytics, companies typically build on top of the machine learning platforms offered by the big cloud providers, Rao said. But well-known enterprise applications are getting decision intelligence capabilities baked right in more and more often.
Dan SimionVice president of AI and analytics, Capgemini
CRM systems or other sales support systems might have decision intelligence software to help sales staff prioritize calls, customize offers and perform other tasks that would typically require a lot of time and effort.
Startups are also entering the space, Rao said. A niche vendor might focus on targeted applications of decision intelligence software, such as making credit card or healthcare recommendations, anti-money laundering or risk analysis.
"It's a very active market," Rao said.
Return on investment depends on the particular use case and implementation and how transformative the decision intelligence becomes for the enterprise.
One common use of decision intelligence software is when banks make loans. Traditionally, a bank officer would analyze an application, perhaps do some research about the specific project or property or ask the applicant for additional information.
Not only is the process time-consuming and subject to personal interpretation, but there are only so many factors that a human can consider.
"What banks are able to do now is have a system which can analyze all the numbers and say, this is a case where you can give the loan, and these are the reasons why," Rao said. "Or it might say that it's a borderline case."
All the routine analyzing and number crunching is handled automatically, and the bank officer can spend more time focusing on borderline cases or dealing with people instead of dealing with spreadsheets.
Decision intelligence and the customer experience
Decision intelligence has been incorporated into all aspects of the customer experience, said Nicole France, vice president and principal analyst at Constellation Research.
Everything from shopping cart and movie recommendation engines and price quoting tools that serve customers directly to routing systems for call centers to support systems for sales and marketing teams can make use of decision intelligence software.
"It's not actually about making the decision for you," France said.
What it does is democratize access to the data and analysis of that data that can help make better decisions. The tricky part isn't pulling the data together or creating the analytics. It's getting the context, interfaces and workflow right so the systems are actually helpful.
There's a great deal of complexity that goes into making something that's clear, easy to understand and simple. "It's difficult to get right," France said.
One area where decision intelligence software can make a big difference for business-to-business enterprise is when it's used to help the sales team make deals.
A decision intelligence framework can take on the administrative work of checking costs, production schedules or conflicting commitments. Or it can tell a salesperson what impact a 5% discount rate would have on their commission or help them decide whether to defer payments.
"You can make those kinds of business tradeoff decisions because you have access to information in a form that's easy to understand and readily accessible," France said.
Vendors are increasingly incorporating decision intelligence functionality into new business applications, she added. It's happening in sales performance management systems, for example. Digital assistants are another way decision intelligence aids this market.
"A lot of salespeople prefer to talk and don't like to type things out," she said. "There are interfaces that are designed to work the way people prefer to work."
For example, a digital assistant might remind a salesperson that a client meeting is coming up in an hour, provide background information on the status of the deal and collect relevant materials from current news sources.
Decision intelligence in manufacturing
Bad decisions at manufacturing companies can cost millions. If a piece of equipment isn't maintained properly, for example, a breakdown can shut down an entire assembly line.
If it's a small operation with a single key piece of equipment and a technician who's been working on the machine for decades, subtle changes in the machine's performance can be the cue that it needs to be serviced. But most manufacturers today have operations too complex to be managed by human instinct, and the amount of information they collect is too vast for human understanding.
In this use case, decision intelligence software needs to involve artificial intelligence and machine learning, said Dan Simion, vice president of AI and analytics at Capgemini.
"Without leveraging AI and [machine learning], it would be nearly impossible to recognize patterns across so many data points," he said.
AI and machine learning are all about recognizing patterns, making predictions and doing so at a speed that can lead to actionable business insights, he said.
"For example, decision intelligence could identify patterns from millions of data points at a very fast pace in order to predict any equipment failures or equipment maintenance for a company's technology systems," Simion said.
Decision intelligence needs to be a key piece of a company's overall analytics strategy, he said.
"Decision intelligence is one of the cornerstones of data-driven insights for enterprises," he said. "You cannot have a data-driven enterprise unless you have intelligent support systems and software."