Picture this: I am a supply chain manager working at a consumer health products manufacturer. I’m excited on what would otherwise be just another Monday morning. Why? Our new AI platform has gone live. We’ve named it Ani. My team can interact with Ani in spoken language to manage our supply chain.
On my way to work, I fire up my smartphone and ask, “Ani, what is my revenue forecast for May?”
Ani answers: “Contract revenue forecast for May is $1.1 billion against a $1.42 billion target.”
I prod: “What regions are below plan?”
Ani: “Contract revenue forecast for May is $1.1 billion against a $1.42 billion target.”
Now that I know the problem, I want to immediately resolve, so I ask Ani: “Is there anything we can do to close the gap?”
Ani responds: “I found excess inventory of Good Teeth 16 oz. in Chicago DC. If it can be shipped to Atlanta DC, an estimated $39 million in backlog can be converted to revenue. Would you like me to execute that transfer?”
I answer: “Yes.”
Ani responds: “OK, I’ll process that transfer. I’ll let you know once it’s complete.”
Wow. This is fantastic.
Not only was Ani able to tell me exactly what is happening in my supply chain, it diagnosed the problem, found a solution and executed the corrective action.
On Tuesday, I want to confirm that the transfer to the Atlanta DC has been processed correctly. Alas, Ani is not as insightful as it was yesterday.
I ask: “Ani, what is my on-hand for Good Teeth 16 oz. at Atlanta DC?”
Ani responds: “Today, in Atlanta, expect mostly sunny skies with a high of 76 degrees and a low of 55 degrees.”
I ask the same question a few more times and get the same, less-than-useful answer. I finally resort to answering my question the old-fashioned way by searching our business intelligence reports.
We have all had similar experiences in our personal lives with Alexa or Siri or our choice of virtual assistants. While this is a frustrating experience, the stakes are typically low. In supply chain, however, the stakes are much higher: revenue lost, delayed shipments or unhappy customers.
The potential, however, is also high. While general AI adoption in the supply chain is still limited, leading companies like Amazon, Procter & Gamble, Schneider Electric and Unilever have already demonstrated the benefits of AI adoption (See The Gartner Supply Chain Top 25 for 2019).
Supply chain digital and analytics leaders can guide their organization’s AI strategy by answering questions like what AI means to the supply chain, what benefits to expect and what challenges to anticipate.
A Common AI Definition
The first step in answering these questions is to establish a common language when speaking about AI. Start with a definition: AI applies advanced analysis and logic-based techniques, including machine learning, to interpret events, support and automate decisions, and take actions.
The State of Adoption
In our research, we found that the top driver for AI adoption is to improve efficiency. AI can certainly help support organizations’ ambitious goals more efficiently by analyzing massive amounts of data and generating more dynamic decisions.
Cost reduction is a distant second driver. That makes sense, since early AI adoption is usually viewed as a major competitive differentiator and reducing costs alone can never generate a competitive advantage.
When we asked companies to tell us the areas where AI is adopted for decision making, 46% of manufacturers and retailers said that they are adopting it in supply chain. This compares to only 34% saying they have adopted it in customer service, R&D and sales. This finding underscores supply chain’s unique position to lead the organization in their AI adoption.
Companies have also indicated that skills availability is their top challenge in adopting AI. The shortage is not only in technical skills for developing AI solutions, but more importantly includes the business and supply chain skills to leverage AI in decision making. This can leverage a strong analytics culture (See: Strategies for Cultivating Supply Chain Analytics Culture).
Data is the second challenge. That means data availability, data relevance, data quality, data sharing, data privacy and data security. Whatever data challenges we have faced prior to AI are only expanding given AI’s full dependence on data to train on identifying patterns, predicting and reaching conclusions.
Tying for second is the challenge of quantifying and understanding the benefits, with 16% of companies citing this as their top priority. We see this challenge front and center among companies expanding beyond pure experimentation, looking to expand pilots to broader deployments.
So, How to Proceed?
When it comes to AI, supply chain organizations face two choices. One choice is to be an early leader, coping with the high risks for high rewards. The other choice is to wait for further adoption by early leaders, to be guided by their experiences before committing the organization’s resources.
As a supply chain digital leader, we recommend you take a multi-step approach to help your organization make the best choice:
- Take the time to educate yourself and your team on AI.
- Look around and get inspired by early adopters’ stories, lessons learned and watch-outs. (See Video: Intel — Autonomous Planning).
- Have a clear vision of why your organization will adopt AI. Make sure this aligns with the company’s priorities for driving efficiencies and digitalization.
- Make sure there is a quantifiable business case for adoption.
- Choose initial adoption use cases that fit your organization’s realities: data availability, technical skills and overall analytics maturity.
- Strengthen your organization’s digital and analytics culture. Their adoption and buy-in will be the true measure for AI success.
In summary, help your organization view AI adoption not as a “do now or die” mandate, but as “a learn fast or be left behind” mantra.
Noha Tohamy, Distinguished VP, Gartner Supply Chain