AI Time to Value: Three Steps to Achieve Results

Last week, Zero100 convened a group of CSCOs in Dublin to discuss how AI is transforming business operations and what senior leaders can do to accelerate the journey. In a series of structured small group discussions, we landed on some takeaways that apply to nearly all kinds of organizations.??

The bottom line: Think big, but also start small and lay the groundwork for instant scale. Portfolios always win and advantage goes to those who balance all three.?

Think Big: AI Is the Real Thing?

General-purpose technology breakthroughs have historically driven radical change in business models and the physical landscape of operations. For instance, steam engines allowed factories to localize energy, freeing England’s early industrialists to build operations away from rivers that powered first-generation mills. More recently, e-commerce has allowed brands to sell directly to consumers, enabling disruptive business models in retail, from Amazon to SHEIN, to reshape factory and fulfillment networks worldwide.?

AI looks like the same kind of transformational technology that created the Industrial Revolution and the Internet Age. As a ubiquitous tool blending statistical analytics and high-powered computing, AI is a learning technology. This means experience breeds improvement. It also means that sitting on the sideline risks missing the boat entirely.??

On the demand side, much of this learning is already mature. Companies from Kimberly-Clark and Starbucks to JD.com and Philips have used AI for years to shape demand, tune the customer journey, and refine pricing strategies. This demand-side AI experience is a key opportunity for supply chain leaders because commercial teams understand it, and financial leaders have seen it deliver results.?

AI is alive and well in many businesses today. Thinking big means envisioning how progress on the demand side can speed up growth with commensurate innovation on the supply side. Closing the supply/demand loop with AI could unlock value with cost savings, customer loyalty, and operational resilience.?

Start Small: Classic AI Counts?

Another important takeaway from our discussions on time to value is that “glaring supply-side business cases” are already visible in most organizations. Top examples include using AI chatbots for long-tail supplier negotiations, as Walmart, Maersk, and adidas exemplify, where cash savings and significant productivity gains are proven.??

Also important are examples that involve using AI tools for predictive maintenance and quality management, as we’ve seen in use cases from Colgate-Palmolive, BMW, and Haier. These cases are established in manufacturing and dramatically better than traditional analog or manual processes. The benefits in labor cost and asset utilization when using classical discriminative AI with machine vision or sensor data are generally an easy sell. Plus, logistics cases based on reinforcement ML applications in route planning and package selection, as UPS, Procter & Gamble, and Amazon show in reality, are, again, proven success stories with simple business cases.?

Examples like these and hundreds of others we’ve collected rarely blow anyone’s mind, but they do validate the usefulness of AI as a general-purpose technology and help paint the bigger picture. Time to value is short, but more importantly, such stories can catalyze transformation on a larger scale.?

Scale Fast: Bring a Lean Mindset to AI??

Speaking of scale, the final piece of the AI three-step is where most leaders see the biggest challenge. Clean, trustworthy data is understood by all as the essential ingredient for any serious AI transformation. Unfortunately, the cost and time needed to build this data foundation have proven to be stumbling blocks. Plus, AI models designed to churn through the data also require constant maintenance and tuning.?

Two takeaways from our discussions resonated as possible solutions. First is to create capacity with the subject matter experts needed to get the data and models straight. Operations people in functional areas know the data as it applies to the actual work, but they are busy doing the work. Freeing them up explicitly is one leadership move that is critical to success.?

Second is the formal use of a continuous improvement process to manage the data and the models. The concept is increasingly formalized in specific job roles and known as AI/MLOps, which some leading companies are already hiring for. The AI revolution is quickly pushing this competence to the forefront for organizations seeking to scale fast.?

AI time to value is on everyone’s mind. A portfolio approach that balances big vision, small wins, and a method for systematic improvement is the way to move faster.?