This article originally published on forbes.com.
FOMO—”fear of missing out”—is driving too much unproductive spinning in many supply chain organizations looking to implement AI. (And they are all looking to implement AI.)
Companies across every industry see AI and machine learning (ML) as key to reaching the “supply chain promised land”—a future state where supply chain disruptions are anticipated and nimbly sidestepped, where complex processes are automated and where happy, empowered teams make smart, data-based decisions in real time.
As of today, however, many chief supply chain officers (CSCOs) are still trying to find their way out of the wilderness.
CSCOs know they want to do something with AI. They’ve all set aside a budget. Many have spun up various councils, experiments and even robust projects to leverage AI in their supply chain operations. However, while many are making real progress, others are lagging behind.
In my conversations with CSCOs, CIOs and other company leaders, I’m finding that the laggards are often being driven by “AI FOMO” rather than a cogent strategy and disciplined approach. Here’s my advice for how to move forward in a way that promises real returns.
“If I were given one hour to save the planet, I would spend 59 minutes defining the problem and one minute resolving it.”
That pearl of wisdom has been attributed to none other than Albert Einstein, and CSCOs would do well to follow a similar approach when formulating AI strategies and programs.
AI has great potential to transform a plethora of supply chain processes—demand forecasting, order management and fulfillment, supply planning, logistics and distribution, sales and operations planning, to name a few. Companies need to be deliberate and specific about the problem they’re trying to solve and weigh the estimated effort versus anticipated returns.
Start with one problem and its associated processes. That said, dive deep into the process, paying particular attention to where the handoffs occur between different internal teams, trading partners and systems. These are where problems often manifest.
Many companies are starting with customer service processes, where the potential positive impact is high and they tend to have plenty of historical data within the four walls of the company.
Diving deep into the process makes it possible to proceed to the next critical step—determining which parts of the process you intend to automate with AI. This is not a trivial step. In fact, it’s hard work to determine what can be automated and what should be automated. What’s technically possible? Where do you need to ensure humans are actively involved or providing oversight? Do you have a plan for redeploying people when time has been freed up due to the automation of certain tasks?
My company’s approach with customers and prospects is to workshop all of this in a dedicated, one-day session. It’s well worth this minimal upfront time investment to get clarity on the problem you’re trying to solve and the path forward.
It’s early innings in the AI implementation game, and let’s not kid ourselves—there’s a lot we don’t know.
That makes it all the more important that companies put their heads together with trusted technology partners. I advise companies to carefully evaluate partners based on three criteria:
1. Rich, varied information sources. AI thrives on high-quality, diverse data. Make sure your chosen AI collaborators have access to a wealth of information across industries.
2. Operational overhaul. Embracing AI isn’t just about new gadgets; it’s about transforming how you work. Look for partners who can adapt to your workflows and business practices.
3. Connected ecosystems. It’s important that your technology stack can play well with others. Prioritize partners who’ve built robust networks that can tap into and share insights across different systems and organizations. This connectivity enables a bird’s-eye view and allows for tweaks at every stage.
I genuinely believe there’s a supply chain promised land on the horizon that most companies will be able to reach, but we have work ahead of us. Part of that includes serious change management to help supply chain professionals navigate this massive transition.
There are no easy answers here. Leadership has to listen to employee concerns and dedicate time and resources to training, upskilling, education and support. Leaders also need to really harness the creativity and energy of employees who are excited about AI because many are. A February 2024 Epicor survey of hundreds of U.S. manufacturing workers found that more than half want to work for tech-driven companies, and more than eight out of 10 felt new tools and tech made them more efficient. My sense is that supply chain professionals are no different.
Lastly, the supply chain industry has to get over its exaggerated anxiety about data sharing and training AI on big, shared datasets. Yes, best-in-class security and data privacy are absolutes, but we can navigate both the technical issues and the competitive/intellectual property concerns.
When I hear hesitation or concern from supply chain professionals, I say: “Look, we have two options. One, we can give you basic, simple output that will not have a lot of intelligence built into it because you’re limiting yourselves to just your data. Or we can give you advanced, intelligent outputs that utilize data from a broader ecosystem. But for you to get it, you have to share your data as well.”
There is a supply chain promised land—and the best and fastest way to get there is together.
