In the latest McKinsey Global Survey on AI we noted a significant year-over-year jump in companies using AI across multiple areas of the business. And while most survey respondents said their companies have gained value from AI, some are attaining greater scale, revenue increases, and cost savings than the rest. Based on our research and experience, this is no accident; how companies build their business strategy, what foundations they put in place, and how they tackle AI adoption in the workplace can all impact their potential for transformation.
Many companies that have spent years developing AI technologies are facing the stark reality that successfully scaling AI requires more than just deploying AI technology. We find that those companies finding more success in scaling efforts are more likely than others to apply a core set of practices. But even these high performers have room for improvement as our research finds not all use the full range of best approaches. So what are some steps leaders can take to get the most out of their AI investments? Here are the six things top companies are doing.
1. They align AI and business strategies.
One fundamental step for scaling AI is bringing together business, analytics, and IT leaders on how AI can create value for the organization and the plan to capture it. AI high performers are two-and-a-half times more likely than others to align their AI strategy with their company’s broader strategy. They are also nearly four times more likely than others to have a clear enterprise-level road map of use cases across business domains. Prioritizing use cases based on feasibility, time investment, and value can help leaders balance short terms needs and long-term value. For example, as business and analytics leaders at one industrial firm laid out their plan for how AI could improve operations, they sequenced implementation of nearly a dozen AI tools over 18 months. They estimate that their structured approach has put them on track to capture tens of millions in profit once all use cases are deployed.
2. They ensure cross-functional collaboration.
AI high performers are almost three times more likely than others to have AI and business experts work together to solve business problems, proving that interdisciplinary teams with diverse perspectives are crucial in AI development. These teams ensure that AI efforts reflect organizational priorities, address end user needs, and achieve value faster. In our research, cross-functional AI execution teams were typically embedded in the business unit for the duration of the project and include a project owner (usually from the business), translator, data scientists, data engineers, data architects, visualization specialist, user interface designer, and business analysts. At one retailer, the use of interdisciplinary teams helped AI experts gain a deeper understanding of how product buyers and store merchandisers performed their jobs. With this insight, they built a more effective model for recommending product placement in stores. Gross margins have increased between four and seven percent in product categories where the tool is applied.
3. They invest in AI talent and training.
AI transformations are as much a cultural change as a technological one. They require new skills, like translator expertise, and new mindsets. This requires upskilling, and all organizations (including AI high performers) will need to invest further in training. Only 35% of respondents from high performing companies and a meager 10% of all others say they have an active continuous learning program on AI for employees.While there are numerous learning programs on the market today, in-house AI training programs (what we call analytics academies) are emerging as a core element for the broad-based learning necessary and critical roles like translators. One retail conglomerate set up a comprehensive training program to prepare its more than 40,000 employees for its AI transformation. More than 1,000 employees participated in the program’s first six months and an additional 150 people go through training each week. This effort helped the company transition from scattered AI use cases and skeptical end users to a portfolio of initiatives anticipated to drive 70% growth in earnings before interest and taxes over the next three years.
4. They empower AI experts with standardized tools, protocols, and methodologies.
Providing AI experts with the right tools and a standard, repeatable playbook can enable them to seamlessly collaborate, fluidly move across the organization to fill talent gaps, deliver real business results more quickly, and better manage enterprise risk. Seventy-six percent of respondents at high-performing companies say their organizations have standard AI toolsets, compared to only 18% of others.Experts need access to state-of-the-art workbenches and tools for managing and structuring data, modeling, and visualizing and simulating data. They should also have repeatable methodologies and protocols for bringing new use cases to production. Companies that do this well follow a structured playbook through every phase of AI development: identifying opportunities, validating the value at stake, evaluating feasibility, developing AI models, modifying how they operate to ensure the models actually capture value, and finally, tracking model performance and updating models over time. AI high performers are nearly four times more likely than others to know how frequently their AI models need updating.
5. They apply strong data practices.
AI high performers are also more likely than others to have a clear data strategy for AI and well-defined governance processes for key data-related decisions. One bank found that improving data management could generate up to $2 billion in annual value, generated from improved cross-selling due to better data, capital savings from reduced operational risks that arise from poor data quality, and costs saving from consolidated data systems and teams. In our work, we find companies are most successful when they have a strong, centralized governance program for data quality and data management. This includes instituting policies about what data can be used; how and where different data sets should be stored; how data quality is monitored and maintained; how data is used and tracked; and how metadata, including data definitions and data lineage, are documented. Successful companies also have a clear data ownership structure, with business units owning business-relevant data and accountable for the quality of the data they generate.
6. They drive adoption and value.
One of the biggest challenges we find all organizations face is simply getting front-line employees to use AI insights in their daily decision making. Our recent research shows the pervasiveness of this issue: just 36% of respondents from high-performing companies, and only eight percent of all others, say their front-line employees use AI insights in real time to enable daily decision making. To make progress here, companies will need to redesign workflows so it’s easy for employees to incorporate AI insights into their day-to-day activities. They’ll also need to empower front-line workers to make data-driven decisions, rather than having to seek their manager’s approval. Leaders should also set in motion a broad set of change management activities to encourage and incentivize workers to use the new tools. For example, the CEO of a retail conglomerate energized participation in their AI transformation by widely publicizing successes across the company and promoting top talent into new roles based on their efforts to develop new AI tools.
By putting these core practices in place, leaders can position their companies to move more quickly and maximize the returns from their AI investments.
This article appeared first in Harvard Business Review.