Driving ROI with AI: Build AI Strategies That Scale

Summary

Three years into the AI revolution, organizations are exploring how to transition from proofs of concept to operational AI that delivers real value. The key question is how to scale AI initiatives, driving ROI and making AI an integral part of business strategy. AI leaders from various sectors—Emin Can Turan, CEO of pebbles.ai; Eleanor Treharne-Jones, CEO of Big Eye; Charlotte Werger, Chief Data and AI Officer at H&M Group; and Ken Moore, Chief Innovation Officer at Mastercard—discussed the essential components of successful AI strategies. They emphasized aligning AI with business objectives, ensuring data quality and governance, and encouraging a culture of innovation and adaptability. Key pitfalls in AI projects include neglecting change management, failing to identify high-value use cases, and inadequate data foundations. The panelists agreed that AI should support existing business strategies rather than dictate them, emphasizing the importance of customized solutions over generic models. Additionally, they highlighted the necessity of trust in AI systems, both in terms of data handling and in building confidence among users and stakeholders. The conversation highlighted the importance of communication and collaboration across departments, encouraging a culture where experimentation and learning are integral to adapting to AI's evolving environment.

Key Takeaways:

• Align AI initiatives with clear business strategies to ensure meaningful outcomes.
• Prioritize data quality and governance to build trust and reliability in AI systems.
• Address change management proactively to facilitate AI adoption across teams.
• Encourage cross-functional collaboration to integrate AI solutions effectively.
• Focus on high-value, low-risk use cases to maximize early AI returns.

Deep Dives

Aligning AI with Business Strategy

AI should not exist in isolation as a separate strategy; rather, it should be smoothly integrated into the business strategy. As Charlotte Werger noted, "There shouldn't be such a thing almost as an AI strategy, but very much a business strategy." This approach ensures that AI initiatives are purpose-driven, addressing specific business challenges and opportunities. The panelists emphasized the importance of understanding the core business problems before designing AI solutions. For instance, Charlotte highlighted that H&M focuses on supply chain efficiencies to deliver significant ROI. This alignment not only optimizes resource allocation but also ensures that AI initiatives have a tangible impact on business performance. Ken Moore added that AI should support the business strategy by enhancing commerce—making it safer, smarter, and more personal—thus reinforcing Mastercard's broader goals.

Data Quality and Governance

Data quality and governance are foundational to the success of AI initiatives. Eleanor Treharne-Jones stressed the importance of reliable data, stating, "AI brings the why now moment" for addressing longstanding data issues. As AI systems increasingly make autonomous decisions, the integrity of the data they rely on becomes crucial. Continuous monitoring and validation are necessary to maintain trust in AI-driven outcomes. Eleanor's focus on building an AI trust platform reflects the industry's shift towards ensuring transparency and accountability in AI operations. This involves not just managing the inputs and outputs but also ensuring that the models themselves are transparent and reliable. In the financial sector, as Ken highlighted, the stakes are even higher, requiring rigorous data handling protocols to prevent breaches of trust.

Change Management and Cultural Adaptation

Effective change management is critical in driving AI adoption. Charlotte Werger pointed out the importance of change management, especially in large-scale implementations like those at H&M, where aligning thousands of demand planners with new AI tools is essential. Ignoring the human element can result in resistance, as Ken Moore observed with Mastercard's AI projects. He noted that addressing employees' fears about job displacement is crucial, shifting the narrative to how AI can enhance their roles. This involves clear communication about the benefits of AI and providing the necessary training and support. The panelists agreed that encouraging an environment where experimentation is welcomed and learning is prioritized can significantly enhance AI adoption and integration.

Trust and Ethical Considerations in AI

Building trust in AI systems is essential, particularly as data-sharing between organizations becomes more prevalent. Ken Moore emphasized that "trust is the currency of innovation," advocating for strong consent models and governance frameworks. This trust extends beyond data handling to include ethical considerations in AI development and deployment. Ensuring that AI systems are designed with ethical principles in mind is essential to gaining stakeholder confidence and avoiding potential misuse. Emin Can Turan discussed the importance of ethical AI, particularly in safeguarding proprietary data and ensuring that AI systems serve the intended business purposes without unintended cross-pollination of sensitive information. This focus on ethics and trust ensures that AI solutions are not only effective but also sustainable and aligned with broader societal values.