8 machine learning benefits for businesses

What is the purpose of machine learning?

A machine learning algorithm examines data for patterns. As the algorithm receives more and more data, it has the potential to improve over time.

The prospect of better predictions dovetails with businesses' data-driven aspirations. For such organizations, ML can deliver recommendations, forecast customer demand and support the corporate decision-making process. The technology has also fueled other AI developments -- most notably generative AI -- poised for enterprise adoption.

Beneficial ways to use machine learning for business

With that backdrop in mind, here are eight leading machine learning benefits for business.

1. Analyze historical data to retain customers

The ability to cultivate customers ranks among the top reasons to deploy ML. Customer churn is a huge headache for enterprises. ML can help businesses identify which customers are likely to leave.

"This is absolutely the No. 1 problem we see with our clients -- whether it's a long-term contract or month-to-month, across different industries and company sizes," said Matt Mead, CTO at SPR, a technology modernization company in Chicago.

We have found that the best return from a financial perspective is where the analytics capability is positioned as close to major revenue sources as possible.

Customer retention is basically a classification problem. Mead said this ML task involves looking at the characteristics of a business' customers -- i.e., historical information on those who have left and those who have stayed, plus their different behaviors. Customers can use that analysis to establish "white-glove programs" for potentially at-risk customers, Mead noted. The business can try to boost customer satisfaction and create a stickier relationship, he added.

David Frigeri, a managing director who leads the East Coast AI strategy at Slalom, a business and technology consulting company, also cited customer retention as an ML benefit.

"We have found that the best return from a financial perspective is where the analytics capability is positioned as close to major revenue sources as possible," he said. "So, building a better customer experience, improving the retention, improving the lifetime value of the customers through better products or services is really the horizontal focus that crosses all the major verticals."

2. Cut unplanned downtime through predictive maintenance

Another in-demand ML application is predictive maintenance for fixed or long-term capital assets, Mead said. Here, ML identifies equipment likely to experience failure. Organizations can use that insight to schedule downtime and make repairs versus experiencing costly outages that disrupt clients, he said.

The global market for predictive maintenance is forecast to reach $19.3 billion by 2028, growing at a compound annual growth rate of 30%, according to Vantage Market Research.

3. Launch recommender systems to grow revenue

Netflix and Amazon offer high-profile examples of using ML to build recommender systems that suggest new products or services based on a customer's purchasing history.

"Those are interesting, very public implementations of ML in the spirit of personalization," Mead noted.

This ML use case creates greater value for customers -- and also opens upselling and cross-selling opportunities for enterprises. A recommender system can thus generate new revenue streams for businesses.

4. Improve planning and forecasting

ML is all about making predictions, so the technology offers a natural platform for planning and forecasting activities.

ML can help businesses predict future costs, demand and price trends to facilitate budgeting and protect a business' financial prospects, Mead said. "That's a huge category of work we do for our customers," he noted.

Within enterprises, the corporate strategist role stands to benefit from greater ML uptake. The trends corporate strategists must consider -- and the pace at which they need to analyze them -- are fundamentally different in light of the COVID-19 pandemic, said David Akers, a research director in Gartner's strategy research group.

AI technologies can lend greater insight and efficiency to the process. But a Gartner study published in July 2023 found only 20% of the 200 corporate strategy leaders surveyed use tools such as ML. Adoption looks set to increase, however, as 51% of respondents said they are investigating ML.

ML's predictive modeling will bolster the foresight necessary for strategic decision-making, helping a business "see around the corners," Akers noted. He cited the importance of unsupervised ML and the ability to "identify new opportunities that we didn't see with traditional analytics."

Unsupervised learning models don't require humans to train data sets and can uncover patterns in unstructured data.

5. Assess patterns to detect fraud

ML and its ability to identify patterns have found a home in fraud detection.

Mead said he sees customers deploy off-the-shelf fraud detection software, but he has also come across a fair amount of custom implementations. Fraud detection is often associated with financial services companies looking for anomalies in credit card transactions.

But Mead cited wider applicability.

"We've worked with customers to identify fraudulent accounts across all sorts of industries," he said. That includes helping e-commerce companies flag fraudulent orders.

6. Address industry needs

While ML has considerable horizontal applicability, organizations can also marshal the technology to meet vertical market requirements. Here is a sampling of industries to consider:

• Financial services. Companies in this sector also benefit from various ML use cases. Capital One, for instance, deploys ML for credit card defense, which the company places in the broader category of anomaly detection. Indeed, the company also uses ML to look for warning signs across its credit card, auto loan and lines of credit businesses.
• Pharmaceuticals. Drug maker Eli Lilly has built AI and ML models to find the best sites for clinical trials and boost the diversity of participants. The models have sharply reduced clinical trial timelines, according to the company.
• Manufacturing. The predictive maintenance use case is prevalent in the manufacturing industry, where an equipment breakdown can lead to expensive production delays. In addition, the computer vision aspect of ML -- one of several emerging technologies in the manufacturing market -- can inspect items coming off a production line for quality control.
• Insurance. ML's use in the insurance industry includes recommendation engines that suggest options for a client based on his or her needs and how other customers have benefited from particular insurance products. Such systems can help advisors zero in on the most relevant offerings for clients and facilitate cross-selling.
• Retail. Computer vision technology plays multiple roles in retail, including loss prevention, personalization, inventory management and planning for the styles and colors of a given fashion line. Demand forecasting is another key use case.

7. Build upon the original investment

Another benefit is the ability to generate multiple returns from an initial ML investment. For example, a retailer that creates a data set to forecast product demand has an opportunity to build upon that investment, Frigeri said. A company might not realize it, however.

"There is this kind of a soft barrier of low expectations around thinking, 'We've got really great demand forecasting -- now we're done,'" he said.

But the data set built for demand forecasting can also help retailers anticipate out-of-stock situations, Frigeri noted. And a retailer that can predict when it will lack a particular product can then build a recommender system for safety stock -- a replacement product it can tap as a just-in-case buffer. Other retailer groups, such as email marketing, can also take advantage of the demand forecast data.

"You actually can get a lot done within that same dollar of investment, but you have to be really thoughtful," Frigeri said.

8. Boost efficiency and cut costs

Automation through ML can trim an enterprise's expenses through labor reduction and improved efficiency.

Customer service is one area likely to see cost savings via machine learning. Gartner estimated conversational AI, which combines ML and natural language processing (NLP), will reduce contact centers' agent labor costs by $80 billion in 2026.

Chatbots, getting an extra push from generative AI, have organizations questioning whether they can start to have fewer call center agents who are on the phone for less time, Mead said.

Replacing call center agents with chatbots is one possibility. But Mead said he views using chatbots to assist human agents and reduce call-handling time as the more creative use of the technology. The idea is to have chatbots listen to conversations, understand the context and assess customer sentiment. That insight, combined with NLP analysis of earlier call transcripts, lets a chatbot provide advice to agents while they are engaged with customers, Mead noted.

Generative AI, meanwhile, opens additional avenues for efficiency, said Zakir Hussain, Americas data leader at consultancy EY. He pointed to a 44% time savings in professional writing tasks and a 55% reduction in programming time, citing research from MIT and Microsoft, respectively.

The emergence of generative AI changes the nature of programming, he said.

"It's not about coding anymore," Hussain said. "We have moved on to an era where it is more about leveraging AI to do the coding, but then wrangling that [output] to make sure what it has generated is actually correct."

In that scenario, Hussain said he foresees many developers becoming "data wranglers."

But automation, while important, shouldn't outrank ML's ability to provide new customer experiences, according to Frigeri.

"Automation has had a huge impact for many organizations in terms of driving productivity, but No. 1 is your customer, first and foremost," he said.