By

Revolutions used to be rare, but it seems as if business is experiencing them more often these days. The latest is machine learning, and some experts are predicting that it will trigger the most sweeping business change since the Industrial Revolution.

Machine learning—also referred to as artificial intelligence, algorithms and artificial mental power—is a term used to describe the science of getting computers to execute tasks and other actions without being explicitly programmed to do so[1].

It’s already part of the average person’s daily routine—it powers optical character recognition (OCR), facial recognition, web searching, Netflix movie recommendations, Amazon product suggestions, junk mail filtering and fraud detection. It is also creating capabilities that have not yet reached the mainstream, such as autonomous cars and neural image caption generators, which are algorithms that can understand photographs well enough to describe them accurately in English.

Its revolutionary power, experts say, is in its potential to optimize business processes. Tasks that take humans days or weeks can be completed in minutes by machines.

A McKinsey & Company graph showing the difference between routine statistical analysis and data generated by machine learning. (Click to enlarge.)

A McKinsey & Company graph showing the difference between routine statistical analysis and data generated by machine learning. (Click to enlarge.)

In manufacturing, this technology can use historical data to improve performance, plan for breakdowns and increase efficiency of maintenance. In other industries, it can determine which job applicants will fit best with company culture, calculate long-term customer loyalty and assess the credit risk of loan applicants.

Considering the exponential growth of available data in recent years (a result of the other revolutions, Big Data and the Internet of Things), machines that learn may become a necessity.

Management expert and author Ram Charan doesn’t see it playing out any other way. In a recent Fortune magazine article, Charan suggests that companies that adopt machine learning (which he refers to as “math houses”) will have a substantial advantage over those that don’t. They will have richer and more direct interactions with customers and be able to collect and use detailed data from those interactions. The data can be converted into algorithms to inform high-level business decisions such as resource allocation, innovation and new product development.

Machine learning is still in the early stages, but there’s no doubt that organizations will rely more heavily on it in the near future. Like Big Data and the Internet of Things, machine learning is intertwined with the business world’s larger move toward digital transformation. Implementing it requires many of the same efforts, including culture change, increased collaboration throughout an organization and a carefully constructed strategy.

In a recent report, McKinsey & Company suggests the following steps for organizations looking to incorporate machine learning:

  1. C-level focus on strategy. Without it, machine learning can become just another tool. “It will provide a useful service,” the report states, “but its long-term value will probably be limited to an endless repetition of ‘cookie cutter’ applications such as models for acquiring, stimulating and retaining customers.”
  2. Think beyond the data science. According to McKinsey, two types of people are necessary for companies to realize value from machine learning: “Quants,” short for quantitative analysts, and “Translators,” who “bridge the disciplines of data, machine learning and decision-making by reframing the quants’ complex results as actionable insights that generalist managers can execute.”
  3. Ensure data is useful and reliable. Identify gaps in the data, determine cost to fill those gaps and encourage cross-departmental cooperation. Appoint a chief data officer.
  4. Start small, with simpler projects, to get buy-in. Share the successes across the company. It will help build support and highlight the behavioral changes that are necessary for machine learning to be effective within an organization.

Machine learning’s promises are big, for sure, but they will not materialize without a significant investment of time and effort. Experts agree, however, that now is the time for organizations to begin formulating a machine learning strategy.

Charan asserts in his Fortune article that companies that clutch to old methods will forfeit advantages like real-time decision-making capabilities, enhanced collaboration and reduced overhead—and will therefore lose out to “digitally minded” competitors.

In a recently published book, “The Attacker’s Advantage,” Charan offers a stronger warning: “Any organization that is not a math house now or is unable to become one soon is already a legacy company.”

Visionary or myopic, thriving or obsolete—what will your organization become?

[1] Stanford University course catalog, Machine Learning.

Leave a Comment

Your email address will not be published. Required fields are marked *