Hardly a day goes by without the announcement of an incredible new frontier in artificial intelligence. From fintech to edtech, what was once fantastically improbable is now a commercial reality. There is no question that big data and AI will bring about important advances in the realm of management, especially as it relates to being able to make better-informed decisions. But certain types of decisions — particularly those related to strategy, innovation and marketing — will likely continue to require a human who can take a holistic view and make a qualitative judgment based on a personal consideration of the context and facts. In fact, to date, there is no AI technology that is fully able to factor in the emotional, human and political context needed to automate decisions.
For example, consider the health care industry, where AI is having a huge impact. Even if AI can support a doctor in making a diagnosis and suggesting medical treatments for a cancer patient, for example, only the doctor herself can factor in the overall health condition and emotional context of the patient (and of the patient’s family) in order to decide whether to proceed with invasive surgery. Most of what we do in health care is not simply in the service of making a diagnosis; health professionals also work with patients to find appropriate treatments that factor in a more holistic, empathic view of their circumstances.
AI technologies can provide managers and employees with accurate data and predictions at their fingertips to support and enable the right decisions in a timely way. But even if an AI system gives employees superpowered intelligence to make decisions, those employees can’t provide real value if their company’s internal bureaucracy requires time-consuming pre-authorization from senior managers before they can act on those decisions. To extract value from AI, employees at all levels of an organization need to be empowered to make final decisions aided by AI, and to act on them quickly. In short, there needs to be a democratization of judgment-based decision-making power.
Much that’s been written about the decision-making impacts of big data and AI has tended to emphasize the importance of having centralized teams staffed with plenty of data scientists. This implies that companies with more data scientists have a better chance of generating business impact. My own experience as a consultant, supported by recent research, indicates a different view: Firms that hire an army of data scientists do not always generate better bottom-line value. Rather, it is the democratization of access to AI tools and of decision-making power among managers and employees that creates more tangible value.
Consider internet platform companies such as Airbnb, where data is at the core of the firm’s business model. Airbnb believes that every employee should have access to its data platform to make informed decisions. This applies to all parts of the organization, from marketing and business development to human resources. For example, employees can monitor in real time how many of Airbnb’s hosts are using the company’s professional photography services and in which location.
Data access is key, but it’s not enough. Employees also need to be given the skills to use and interpret that data. For Airbnb, it’s simply not possible to have a data scientist in every room, and the fast internationalization of the company makes the situation even more challenging. So in 2016 Airbnb launched its own Data University, which now has a curriculum of more than 30 modules. The goal is to build employees’ knowledge and skills so that they can utilize and interpret data and its associated tools. This will enable employees to act swiftly on innovation opportunities. For example, product managers are learning to write their own SQL code and interpret their own experiments about whether to launch a new product feature in a certain city. The result: Since launching the program, more than 2,000 employees have been trained, and the weekly active users of its internal data platform — a proxy of how “data informed” the organization is — rose from 30% to 45%.
Another case is Unilever. Orchestrated by the company’s newly created “Insights Engine,” Unilever introduced a number of AI-driven systems and tools that are accessible to all of its global marketers. The availability of real-time, frequent, data-driven consumer insights has generated even more need for distributed decision-making by the company’s marketers. One tool they use is People World, an AI platform able to mine thousands of consumer research documents and social media data. The platform is able to answer natural language questions that marketers may ask on a specific area. This has helped to address the problem summed up in the phrase, “If only Unilever knew what Unilever knows.” It has also helped to remove silos, increase employees’ trust in “one consolidated source of truth” and dramatically reduce the time needed to make informed decisions.
Over the last decade, the costs and time associated with organizing data and running analyses has dropped dramatically. But in many companies, AI use is still highly centralized. Corporate AI units often develop dashboards for senior executives, which are used exclusively by them. AI democratization remains limited. But by using AI to increase the effectiveness of the decisions employees are making, the need to control and centralize decisions essentially evaporates. Best practices show how democratization can bring about quicker and better distributed decisions, making companies more agile and responsive to market changes and opportunities.
