Legacy Companies’ Biggest AI Challenge Often Isn’t What You Might Think
When starting out to deploy artificial intelligence (AI) and machine learning (ML), executives of legacy companies often view the challenges mainly as technical problems — particularly finding sources of internal data to analyze and choosing the right tools. What they may not appreciate is just how data-rich their legacy companies already are.
From utilities and mining, transportation and shipping, to financial services and more, legacy company operations and customer interactions generate a wealth of data. Such data can be harnessed to tackle a very wide range of issues: optimizing supply chains, predicting maintenance, reducing accidents, increasing production output, improving operational efficiency, raising revenue productivity, and growing customer value.
To realize these opportunities using AI, however, legacy companies worldwide typically soon discover that their biggest problem is not technology — it’s talent. Demand for data scientists and analysts is intense and continues to exceed supply. Amazon, Facebook, Google, and other tech leaders hire massive numbers of data scientists, offering them fascinating challenges and compelling opportunities. By comparison, from the viewpoint of a sharp data scientist with leading-edge AI proficiency, a 100-year-old company that makes tractors, manufactures appliances, operates power plants, or ships containers may seem “boring.”
In addition, legacy companies are often located outside of major tech hubs such as Silicon Valley, Seattle, Austin, New York, or Los Angeles — all of which can make it even more difficult for legacy companies to find the data scientists they need. There is a solution: a two-pronged talent strategy of hiring externally and building internally.
Recruiting Talent Using Interesting Problems
To attract data scientists, legacy companies can and should focus on the compelling, unique, and real-world business problems that they offer. As Grant Case, director of sales engineering for Dataiku, a leader in applying AI and ML for enterprises, who works with legacy companies in Australia and New Zealand, told me recently, “We need to give data scientists interesting problems to work on and turn into value. That’s where the magic happens.”
Virtually every legacy company across all industries has very complex — and thus very interesting — questions and problems that offer robust opportunities for intellectually curious data scientists to dig into, such as:
● Unsnarling extraordinarily complex airline systems when weather closes multiple hubs
● Optimizing electricity grids and storage in a world of distributed, multi-directional, production, transmission, and storage
● Predicting accidents to reduce on-the-job injuries
● Optimizing global shipping networks and supply chains in real time for millions of containers every day
● Maximizing crop production from each square foot/meter of earth
Berian James, head of data science and AI at Maersk, the global shipping giant, described optimizing their shipping network as “a really interesting data science problem.” Maersk uses AI and ML to address a wide range of problems and opportunities, from providing its customers with “arrival intelligence” for their shipments to advancing the company’s decarbonization efforts.
Virtually every legacy enterprise, if executives stop and think about it, offers fascinating business questions, problems, and challenges that can stimulate the intellectual curiosity and challenge the technical proficiency of data scientists and AI talent. Thus, an emerging best practice for legacy companies to recruit the talent they need is to use these interesting questions to offer data scientists fresh opportunities to personally address and have an impact in solving engaging, unique business problems. Such scenarios may be more appealing than becoming the latest addition to the multitude at Facebook, Apple, Netflix, Alphabet, and similar firms.
Developing Homegrown Talent—Combining the Right Aptitude with Business Understanding
Hiring data scientists externally isn’t the only solution. While it’s not the answer in every case, developing data science and AI proficiency with internal talent is often faster, easier and more productive, and can be more than sufficient for a wide range of business purposes. Internal subject-matter experts, who have the right aptitudes and interests, already understand the business. This can be more desirable and impactful than going outside the company to hire a data scientist who — although technically advanced — is unfamiliar with the industry and business-specific or company-specific problems and challenges. I’ve heard many stories from executives at legacy companies that hired data scientists and embedded them into the business with great hopes — only to be disappointed when it proved difficult to integrate those data scientists with the ongoing business management and processes.
While internally developed talent may not replace the most advanced data scientists for the knottiest problems, they can often significantly advance the company’s AI and ML use and produce material business value. Certain disciplines found within legacy companies are particularly well-suited to developing AI and ML expertise. Engineers of all types, operations researchers, physical scientists, revenue managers, and others typically have the technical foundation, quantitative aptitude, proficiency with data, and intellectual curiosity to learn how to apply AI and ML and develop the capabilities to do so.
Case gave the example of a steel company where chemists and metallurgists deal with production challenges that could be addressed with data and AI. “You can find talented individuals who want to progress in their careers and enable them with the right training,” he told me. Plus, they typically have the important advantage of understanding the business and, thus, credibility with business leaders.
Solving the People Problem
It is increasingly evident, in talking with executives in a wide range of legacy companies who are working to apply AI and ML, that the biggest challenges are culture, connecting data science and AI to business management and processes and, particularly, finding the talent needed. It’s not primarily a technical problem. As executives of these companies tell me, the ongoing challenges are finding the right people and incorporating them, along with AI applications, into the actual working of an enterprise.
These observations demonstrate that now, more than ever, using data science and AI to realize practical gains requires adept business leadership. Senior leaders must understand what really drives and enables data scientists so that their companies can attract, grow, and integrate this talent in a legacy business to create business value.