Myth vs. Fact: Four Realities of Applying AI to Decision Making
Advances in AI are reducing complexity in executing decisions across the enterprise and revealing new value from decision automation.
Consider a global consumer packaged goods leader that started the decision intelligence journey focused on automating forecasting decisions in the U.S., its largest and most complex market. The technology began quickly learning the patterns of the demand at each state level, producing initial results within a few months. The company has reduced forecast error from 68% to 25% and is scaling now to service areas in EMEA, APAC, and LATAM, with an eye on improving supply planning in the near future.
While early adopters are paving the way for others to take the first step, misconceptions about AI-powered decision making exist.
Let’s start by addressing four myths.
Myth #1: Decision intelligence is only a concept — yet to be implemented
Fact: Decision intelligence software — defined by IDC as packaged for fully or partially automating all steps in the decision-making process — is actively improving cost, revenue, service, sustainability, and more.
According to IDC research*, AI-powered decisions drove up to 20% improvement across a range of business metrics for enterprise “leaders.” In the study, “leaders” were defined as those successfully operationalizing AI for better decisions through use of decision intelligence solutions.
The technology is enabling teams to understand, evaluate, and improve their decisioning processes. They can capture knowledge about how decisions are being made and ensure consistency and continual learning — all factors that deliver tangible value.
Myth #2: Decision intelligence requires perfect data
Fact: Perfect data is an enterprise myth that many have been chasing for years. The reality is, leaders and their teams have to make decisions with incomplete information at speed.
This is where AI applications like decision intelligence help bridge the gap.
Through a data model that integrates all available data in a harmonized layer, users can apply machine learning and decision intelligence techniques to predict missing or incomplete data — gaining a starting point to make informed decisions. By collecting data about expected and actual outcomes, leaders can also understand if their data and the decision logic they are working with is delivering the intended results.
This ability to work with incomplete data and advanced logic at speed and scale will differentiate the winners from the rest.
Myth #3: Decision intelligence will result in loss of control over decisions
Fact: Trusting AI to make and execute decisions requires users to be engaged in the process. It requires explainable AI and a transparent user experience, with full visibility into the data and intelligence behind every recommendation and decision.
In contrast to a “black box” where workings aren’t clear, decision intelligence technology can provide transparency and be auditable, allowing users to understand the basis for recommendations.
This type of intelligent human/machine collaboration drives confidence in the process and enables adoption and outcomes. Whether users are reviewing recommendations and then manually executing the decisions (human-on-the-loop) or identifying groups of routine or specific decisions that can be fully automated (human-out-of-the loop), teams must be able to see the rationale and context for the recommendations (glass box). This will build the trust needed between people and machines to innovate, scale, and maximize results.
Myth #4: Decision intelligence eliminates jobs
Fact: Automating routine work and enabling time for other priorities are just a few workforce benefits of decision intelligence. Its application is also proving a driver in attracting and retaining talent — people want to work with intuitive technology that speeds results and delivers efficiencies.
One global CPG leader changed its organizational structure to accelerate a digital supply chain — prioritizing decision intelligence in this evolution. The critical key to success? The company created the right culture for testing, learning, and adoption. Employees were motivated by clear goals and celebrated milestones as AI-generated recommendations were accepted and higher percentages of automated decisions were realized.
Another supply chain executive leading the technology’s enterprise-wide scaling is now defining decision-specific roles and responsibilities for her team, following the company’s initial test nine months ago.
We are already seeing decision architects, decision engineers, and other decision-centric roles emerge as companies prepare for the future of decision making.
Adopt a decision-centric mindset for the future
AI-led decision-centric initiatives are enabling companies to empower their workforce and communities, deliver time-to-value and value-over-time, and compete effectively in our digital world.
The time to embrace this revolution is now.
A version of this article was previously published on the Nasdaq blog.
Hear from companies that are pioneering AI for decision making. Join us at AeraHUB 24, the Decision Intelligence summit.
*Source: IDC White Paper, commissioned by Aera Technology, What Every Executive Needs to Know About AI-Powered Decision Intelligence, IDC #US51338623, November 2023
