Generative AI will first be successfully scaled in business operations

Generative AI (gen AI) had an exciting year in 2023. This year, claims of its transformative potential will be tested as organizations attempt to scale gen-AI-powered activities. The aim will be to make gen AI part of the fabric and architecture of business operations in a way that measurably moves the dial on business performance. We estimate that gen AI could offer savings opportunities of $1.4 trillion to $2.6 trillion across operations functions, including customer service, R&D, manufacturing, supply chain, and procurement, alongside its impact on the back office.

This will be harder than much of the coverage of gen AI—so far—may have led some to believe. Amid the declarations and promises, we offer some practical ideas for putting gen AI to work in your business.

AI’s potential spans the 4Cs

Gen AI applications span several archetypes of capabilities that reside, at least partially, in the scope of operations functions. These include the 4Cs:

  • Concision. New capabilities in concision have equipped gen AI to interpret large corpuses of unstructured data to identify and summarize relevant answers in service and analysis contexts.
  • Creative content. Gen AI’s potential handling of creative content can enable the rapid tailoring of complex and structured documents to specific needs and contexts.
  • Customer engagement. Out-of-the-box copilots powered by gen AI can guide customers through their personalized journeys in the realm of customer engagement.
  • Coding and software. New capabilities in coding and software promise swifter migration from legacy systems at scale.

Persuasive examples already exist. The customer support function of a South American telecommunications company used conversational AI to prioritize its higher-value clients while promoting self-service. By automating a proportion of its contact activity and consolidating redundant platforms, the company reduced operational expenditures by roughly $80 million. Elsewhere, a gen-AI-powered learning platform led to onboarding surveys reporting improved onboarding experience by some 35 percent. Another business reduced financial planning and analysis costs by more than $6 million through use of a sophisticated gen-AI-powered research assistant that automatically pulls information from multiple sources, synthesizes knowledge, and presents it for human verification.

Operations: The realm of tangible testing

Operations functions are an attractive area for introducing gen AI, because these functions typically have well-established measurement and reporting processes, which make it easier to see the impact of decisions such as how much time a supervisor saves or how much more efficient a particular stage of process has become. Smart businesses will experiment with gen AI in operations, analyze the results, and then carefully apply what they have learned to more complex scenarios. Despite some of the lofty claims made in 2023, most businesses will experience no silver bullet or lightning strike but instead testing, learning, and iterative progress.

For example, at manufacturing plants where shift reports are routinely handed over between shifts, gen AI has the potential to reduce delivery time for these reports by 50 to 70 percent. Organizations can apply gen AI to such workaday but business-performance-enhancing tasks. Those that do so will have live case studies from which to learn and on which to build. Then they can apply the resulting knowledge and know-how to refinements in inventory, scheduling, and the use of raw materials.

Avoiding ‘pilot purgatory’

As in past digital transformations, the best practices for introducing generative AI will involve setting up governance structures; drafting, updating, and socializing transformation road maps; and establishing an indefatigable communications strategy. Creating value from gen AI requires tackling operational readiness challenges as much as grappling with new technologies. This is the familiar terrain of capability building and change management: developing new capabilities in IT and tech, managing risk and reputation, and monitoring regulatory matters. It is critical here to have a strong relationship between operations leaders and tech leaders, as is true for any successful change program.

Companies can benefit from addressing the deployment of gen AI as a transformation, not merely a technological advance. This calls for focusing on the business challenge, not the technology itself. In other words, companies identify the exact business challenge gen AI could address and then verify that a more efficient solution cannot already come from traditional AI, internal rules, or organizational shifts. Deploying gen AI for its own sake will not yield tangible business results and could even become a fruitless distraction.

Building the right team

Quick-win use cases deliver value and excitement, and they prevent efforts from becoming “just another IT project.” Lighthouse use cases foster trust and alleviate organizational concerns while paving the way for more advanced gen AI applications. They also provide the business with a secure space in which to learn and formulate the right questions.

A core team with the right complementary skill set to steer gen AI pilot projects should have expertise in business operations, technology, and change management. Again, the key stages will be familiar to many: clarifying stakeholders’ roles and responsibilities, identifying and elevating domain experts. And as with any major change in ways of working, the chances of success are vastly greater when project leaders involve the front line early and often. When projects are derailed, the most common reason is by a failure of the project leaders to take people with them, and gen-AI-powered projects will be no different.

Involving the right people

Any venture that aims to scale gen AI will involve legal, privacy, and governance issues. Those responsible for addressing these issues need to be on board, and the company should tap their expertise to inform the road map for scaling gen AI. A significant introduction of gen AI is likely to require new controls, training modules, and more. For more examples of current and possible applications for GenAI, and the inherent risks, listen to our recent podcast here.

When it comes to talent, most organizations will likely benefit from upskilling existing tech roles to include emerging gen AI skills, such as prompt engineering. Developing separate roles may be less of a priority, though external hiring in key areas may be necessary. Knowing when to hire and when to train internally for gen AI success is a value creating decision-making skill that leaders will need to master at pace.

Familiar questions

Though companies’ answers will differ, the business questions remain the same: How will a gen AI transformation get us to market faster or enhance productivity and efficiency? What new set of capabilities do we need within the workforce to make the most of the opportunity? How do we measure gen AI’s return on investment?

As companies begin experimenting with use cases, answers to these questions will begin to emerge in the very near future, and many of the success measures already exist. Over the coming years, we will see whether the excitement in 2023 was overdone or gen AI becomes a critical, game-changing tool of the magnitude of, say, data analytics. In the meantime, businesses and their leaders have work to do and choices to make as they test ideas and search for value through the smart application of these new technologies.

Connect with our Operations Practice