The Signal June 11, 2024

Zero100 AI ROI: 5 Steps to Take You from Science Experiments to Tangible Results 

How do you unlock AI ROI? Not with an AI science fair. At a Zero100 event last week, supply chain leaders shared how they’re cutting through the noise and breaking down silos to deliver a real impact end to end. Here are our five key takeaways.

Greta O'Marah Avatar
Greta O'Marah

At a science fair, the key ingredient – for erupting volcanoes, building bottle rockets, and making invisible ink – is baking soda. In supply chains, a key ingredient – for automation, simulation, productivity, and much more – is rapidly becoming AI. Last week, Zero100 brought together a small group of supply chain professionals for an interactive cross-industry learning session focused on moving from AI experimentation (ahem, a “science fair”) into scaled AI implementations, tangible ROI, and accelerated time-to-value. 

In the 68 years since Stanford professor John McCarthy defined AI as “the ability for a machine to complete thinking tasks once thought only achievable by humans,” its business applications have proliferated thousands of times over. And in supply chain, 2024 marks a pivotal year for AI adoption as CSCOs, COOs, and their digital and strategy leaders are tasked with delivering measurable results.  

It’s likely you can’t go a day without a new promise for an AI savior for your supply chain landing in your inbox. With so much hype, it can be hard to spot the real opportunities. In 2024, 44% of earnings calls mentioned AI/ML, but of that 44%, only 25% are seeing AI drive revenue right now, suggesting a mismatch between vision and the reality of unlocking value.  

Bar chart showing percentage of earning calls in which AI/ML is mentioned (2021-2024) and ring charts showing percentage of companies who have an AI 2024 vision vs those for whom AI/ML is currently driving revenue.
Source: Zero100 analysis of earnings calls.

We dove into what’s possible today with AI and how it can accelerate the path to value at our AI ROI event last week in Chicago. Our focused group of 30 cross-industry supply chain leaders landed on five takeaways: 

1. A picture is worth a thousand words… visualize AI across your end-to-end supply chain.   
Zero100 data shows that leaders in digital business, like the Estée Lauder Companies, General Mills, and AstraZeneca, are moving fast on integrating AI pilots across their supply chains compared to those experimenting in silos. Across 100 global companies, the top quartile of AI performers grew revenue nearly twice as fast as the bottom quartile group, and they did so at +2.88% higher gross profit. The key is to evaluate opportunities, weigh effort vs upside, and then fail fast. Our AI blueprinting process is built to help organizations accomplish exactly this.  

2. Prioritize initiatives… for practical wins.  
What are the still untapped general AI opportunities in supply chain, both at a functional level and across end-to-end operations? How do you make the right bets among the many options? We got into some answers at the event, looking to real-life examples like Shell, who bet on sourcing, and invested in the transformation of its Source-to-Contract (S2C) and End-to-End (E2E) processes. It now measures 50 key data points and 4,000 clause elements to draft contracts for supplier bids, and after deals are executed, RPA integrates key information into Shell's ERP system, reducing cycle time by 25%. 

3. Manage the data generated and consumed by AI… to allow for sustainable growth. 
In this data revolution, data management is key. And it's all about finding the right people. The most digitally advanced companies, like Walmart, Microsoft, Toyota, and Procter & Gamble, seek out a mix of analytical abilities and emotional intelligence when recruiting talent, helping to drive technology-enabled innovation in supply chain processes.  

Newer capabilities like MLOps, retrieval augmented generation (RAG), and reinforcement learning account for the fact that AI/ML models require continuous, collaborative management, which allows organizations to keep pace with the dynamic business environment. Despite this, as of March 2024, only 14% of the major brands we surveyed recruited specifically for MLOps – a capability with the power to transform long-term data management. 

4. Get the board on board… by crystallizing a bold vision and perfecting your pitch.  
Boards are now regularly asking for details on the big picture AI opportunity for supply chain. Amazon’s PR/FAQ (Press Release/Frequently Asked Questions) is one approach to clarifying goals, with teams defining output goals in the form of a short and sweet press release that also answers internal questions. It looks ahead to the future, meaning it is unconstrained, but is still grounded in the language of the customer.  

5. Connect the audacious goal to a series of pilots… to set a precedent.  
A participant in the room shared that they’re reaching AI time to value in less than eight weeks on select pilots. Achieving quick time to value on some of the smaller projects might be the fastest path to the big bang delivery of your bold vision. Start by unleashing small, focused experiments, ranging from hackathons to targeted pilots, to test hypotheses, learn from them, and stairstep your way toward the upper right of the chart below. This feedback loop can drive expedited results. 

Today, both mob fears and enthusiasm for AI are peaking, making it difficult to know the best route to unlock AI ROI. In our latest report, inspired by Warren Buffet’s bold bets and backed by Zero100 research, we delve deeper into where to invest to see the most – and fastest – AI ROI.