Supply Chain in the GenAI Era: from Vulnerability to Resilience

With nine out of ten supply chain leaders reporting challenges in 2024, supply chain disruptions have clearly become the norm rather than the exception.

Despite this reality, organizations still face significant gaps in their ability to identify and mitigate supply chain risks, and few have launched initiatives to address these weaknesses.
‍

In our recent webinar with McKinsey & Company, we explored findings from their 2024 Global Supply Chain Leader Survey to provide supply chain leaders with both a clear view of the current situation and insights into building resilience.

This digest from our December 4th, 2024 webinar offers fresh perspectives on three key topics:

• The impact of recent black swan events on Supply Chains.
• Progress made in Supply Chain risk management and resilience, and why vulnerabilities persist.
• Which areas require further actions to mitigate the risks and what opportunities GenAI offers.

Note that can request to receive the complete replay via email at marketing[at]pelico.io.

‍

Meet the Webinar Team
‍

‍

‍

‍

‍

‍

‍

‍

‍

‍

The Impact of Recent Black Swan Events on Supply Chains

‍

Black Swans are here to stay

‍

‍

Over the last few years, we've all witnessed numerous disruptions in our supply chains. From floods and earthquakes to Houthi rebels attacking vessels and the recent ship collision with Baltimore's bridge—these are just some examples of what we call "Black Swan events."

‍

We call them Black Swan events because they are unpredictable and have massive impact. This creates a unique challenge for supply chains: while we're used to dealing with unclear demand and supply fluctuations, these disruptive events are in a different league altogether. Their sheer magnitude makes them exceptionally difficult to manage.

‍

Dr. Alicke's clients at McKinsey & Company often ask, "When and where will the next Black Swan event occur?" However, by their very nature, Black Swan events cannot be predicted. Only one thing is certain: when they occur, their impact is substantial.
‍

In conversations with clients, they often ask, "When and where will the next Black Swan event occur?" The reality is that, by their very nature, we can't predict these events. What we do know is that when they happen, their impact is substantial.

Dr Knut Alicke
‍Partner at McKinsey & Company
‍

Note: McKinsey and Company's teams use the term "grey rhino" for more predictable major events. Trade and tariff discussions, for instance, fall under the category of grey rhinos.

‍

‍

‍

‍

Financial Impact of Black Swan Events

‍

To assess the financial impact of Black Swan events, McKinsey & Company conducted extensive research. While some may dismiss disruptions as routine occurrences, McKinsey's findings revealed striking data:

• The frequency of severe disruptions has risen dramatically in recent years.
• Disruptions put 45% of a company's decade-long EBITDA at risk.
• A typical disruption lasts about two months.
  ‍

‍

‍

‍

‍

‍

‍

‍

The Cascading Operational Impacts and Costs of Black Swan Events

‍

‍

Recently, Boeing experienced their first strike in years. This Black Swan event lasted seven weeks and added $1.1 billion to Boeing's wage bill over four years. This represents only Boeing's direct costs—imagine the cascading effects throughout the entire supply chain network.

‍

Similar to Boeing, Pelico's customers encounter frequent disruptions. Based on data from the factories it equips, Pelico observes that customers face a disruption every 16 minutes on average—whether in Europe, Asia, or North America. Though each customer starts with an apparently ideal setup—including sophisticated ERP systems and meticulous production plans—challenges invariably emerge and ripple through the system shortly after daily operations commence.

‍

‍

Visible Yet Stagnating Progress on Supply Chain Risk and Resilience
‍

‍

Each year since 2020, McKinsey & Company conducts a survey on Supply Chain Risk, asking their clients—heads of supply chain, heads of planning, chief supply chain officers, and COOs—three main questions:

1. Did you face any disruption?
2. What did you do?
3. What do you want to do?

‍

Here’s the overview of the findings from the 2024 McKinsey Global Supply Chain Leader Survey that Dr. Knut Alicke provided during the webinar.

• ‍2020 | post Covid-19 Pandemic: Supply Chain leaders plan many resilience actions, including war rooms. Their goal: ensuring a clear visibility on what their supply chains could deliver.
• ‍2021 | Moving beyond firefighting: Supply Chain decision-makers begin planning longer-term solutions like digital transformations and strategic changes, while maintaining short-term buffers during implementation.
• ‍2022 | First resilience initiatives go live: Initiatives launched the previous year take shape, and many Advanced Planning Systems (APS) are implemented.
• ‍2023 | Cash-flow over Supply Chain risk mitigation: Due to the global economic context, organizations face intense pressure on cash flow, leading them to shift focus away from supply chain risk resilience.
• ‍2024 | Past initiatives show progress, but no new initiatives are launched: Organizations that launched their first initiatives are seeing progress, but the focus on cash flow management means no new initiatives are being launched.

‍

‍In a nutshell:

On the bright side of things: the survey confirms that an increasing number of companies implement formal planning systems, hence improving visibility and increasing transparency.

On the not so bright side of things: companies are not where they’d like to be and still have a lot more to do.

‍

‍

2024 McKinsey & Company’s Findings on Supply Chain Risk

‍

‍

‍

‍

What’s Improving

‍

• Diversification of mitigation strategies:
  For the past three to four years, dual sourcing and regionalization strategies have dominated supply chain discussions. These approaches continue to gain traction, with 60% of surveyed companies implementing dual sourcing and 73% pursuing regionalization as of 2024.
  ‍
• Inventory buffers decrease for the first time since 2020:
  While implementation was initially slow due to the time-intensive nature of these measures, we're now seeing accelerated adoption rates. Inventory buffers are beginning to decrease, and various risk and resilience measures are taking effect.
  ‍
• Advanced Planning System (APS) implementation:
  ‍Planning initiatives are yielding promising results. 68% of companies are implementing Advanced Planning Systems (APS), which also serve as digitization efforts. While these projects take considerable time, they demonstrate significant forward momentum.
  ‍‍‍
• Tier-1 visibility has improved by 10%p since 2023
  ‍‍
• ‍Enhanced internal risk capabilities:
  ‍76% of companies surveyed report having sufficient in-house risk management capabilities, supported by dedicated organizational structures wether centralized or de-centralized.

‍

What’s Has Yet to Be Improved
‍

• ‍Insufficient Supply Chain Risk Understanding:
  Despite progress made, many crucial investments remain behind schedule. Consider supplier visibility—while we have good insight into our tier-one suppliers, visibility beyond tier two is severely limited. 70% of boards lack a complete understanding of supply chain risks' depth and complexity.
  ‍‍
• Early warning systems: Organizations have underinvested in early warning systems that could bridge the gap between tier visibility and planning. While their IT infrastructure performs well under stable conditions, it struggles during disruptions. This explains why identifying and mitigating issues remains a longer process than it should.
  This challenge stems largely from the complexity of connecting and integrating internal and external data sources into our planning systems to generate actionable early warning signals.
  This affects organizations’ ability to identify impacted customers and determine mitigation options.
  ‍

Most IT infrastructure work well during normal operations, but when disruptions occur, organizations struggle to quickly identify issues and understand their impact on the supply chain.

Vera Trautwein
Senior Supply Chain Expert
McKinsey & Company
‍

‍

Risk Mitigation

• Only 48% of organizations upgrade end-to-end planning to model risk scenarios, while 74% focus on machine learning forecasting.
• Companies take an average of 2 weeks to respond to new disruptions.
• 90% of surveyed companies report lacking in-house digital talent.
  ‍

AI adoption in Supply Chain

While we're seeing initial successes, these are primarily limited to demand planning and mid-term planning. The survey shows that supply chain leaders haven't yet leveraged AI to help manage Black Swan events or other major disruptions.

‍

‍

Resilience Through Real-Time Synchronized Supply Chain Operations

‍

Over the past few years, companies have operated under the assumption that everything will follow their plans, leading them to invest in lengthy IT projects spanning three years. Now, a mindset shift is taking place as companies recognize that reality rarely aligns with these plans.
‍

Companies know that things will not work according to plan, but the systems they’ve in place can only work in perfect conditions. They cannot react to the disruptions they see on the day to day.

Henrique Valer
VP Customer Operations and Success North America
Pelico
‍

‍

‍

The core problem lies in their systems' design—they only function under ideal conditions and cannot adapt to handle daily disruptions.

Another major challenge with existing tools is their disconnected nature, which restricts data access to technical experts only.
‍

There's a lot of buzz around democratizing data access, but this goal remains unmet. While technical users can work with the data effectively, it remains inaccessible to the broader organization.

Henrique Valer
VP Customer Operations and Success North America
Pelico
‍

While making data accessible throughout organizations has become a key priority, operational teams remain largely disconnected. This creates a significant gap between processes and, more crucially, between the teams and departments responsible for managing disruptions.

‍

‍

‍

Case Study: Leveraging a Supply Chain Operations Management Platform to Synchronize Operations Seamlessly

‍

‍

‍

Let's explore how a supply chain operations management platform can improve performance through the example of one of Pelico's aerospace customers: one of the largest providers of aerospace systems.
‍

About this Customer

• Multiple high-revenue factories
• Revenue per plant: $500 million to $1 billion
• Location: USA
  ‍

Before Pelico

Despite being one of the most outstanding organizations in their field, this company would have been seven weeks behind schedule if they relied solely on their ERP plans, with all sites in recovery mode.

‍

With Pelico

With Pelico's Production Control App, their production controllers gained powerful tools for automatic shortage calculations, production planning adjustments, and real-time "what-if" scenario simulations. The impact was remarkable—within three months, they transformed from being seven weeks behind to anticipating shortages two to three weeks in advance.

Spotlight on Supply Planning and Advanced Analytics Use Cases

‍

‍

‍

When COVID-19 broke out, most organizations focused on supply planning, investing heavily in supply planning and end-to-end planning capabilities. However, they quickly realized that building resilience would require more than these initiatives. From that point, demand planning began to take priority.

‍

Initially, when COVID-19 broke out, everyone was frantic about supply planning and tried to invest in supply planning and end-to-end planning capabilities. However, they then realized: “hum, this is more complex than we thought.
‍
Vera Trautwein
Senior Supply Chain Expert
McKinsey & Company
‍

Fast forward to today, with demand planning being the number one planning within the  where advanced analytics and other digital use cases have been implemented. Supply planning, meanwhile, took a bit of a back seat.

‍

Supply planning—particularly end-to-end functionalities like inventory management—remains a key priority. Companies continue to work toward better visibility and transparency in identifying and responding to risks.

‍

The survey highlights worrying trends: companies are underinvesting in both early warning systems and complex supply planning capabilities. These gaps make supply chains vulnerable to future Black Swan events and explain why organizations take two to four weeks—sometimes longer—to respond to disruptions. Without fully connected planning data across the end-to-end supply network, reaction times inevitably remain lengthy.
‍

‍
The thing that really worries us is the low investment in early warning systems and the more complex supply planning topics.This is one of the reasons we believe supply chains are still not ready to respond to a new Black Swan event. It’s also why it still takes two to four weeks, or even longer than a month, to react to disruptions.

Vera Trautwein
Senior Supply Chain Expert
McKinsey & Company
‍

‍

Our call to action is clear: supply chain leaders must shift focus from the proven demand planning use cases toward more complex initiatives in supply planning and early warning systems.

Regarding end-to-end planning and APS system deployments: while two-thirds of companies have embarked on this journey, many implementations are falling behind schedule and stalling in the deployment phase.
‍

‍

Key Takeway

‍

Prioritize high-impact use cases that will strengthen your position for the next Black Swan event. This approach should accelerate your APS deployment while bridging these two critical areas.

‍

‍

Pelico’s Vision: Putting Your Supply Chain Operations on Auto-Pilot

‍

‍

Pelico's vision of supply chain operations parallels the evolution of self-driving cars. The journey started with paper maps, evolved to digital maps and GPS, and transformed when consumers and producers began interacting with the system. This interaction enabled navigation systems with guided routing, ultimately paving the way for self-driving cars through the vast amounts of data these platforms generate.

‍

Pelico's evolution mirrors the path that led to self-driving cars. Companies initially managed frequent supply chain disruptions using spreadsheets and homegrown tools. Then came ERPs, though only for some organizations. From these early ERPs, we progressed toward synchronization, achieving functional visibility and collaboration, though still requiring significant manual effort.

‍

Operations Co-pilot marks the beginning of human-machine symbiosis. While humans continue to make high-level decisions, these shift from tactical to business-focused choices, supported by simulations and optimizations.

‍

Finally, Pelico's full potential emerges in what we call autopilot. Like a self-driving car, it enables you to establish high-level rules and let the system guide you toward your goals. This capability is increasingly enhanced by AI and generative AI.

‍

‍

How to Ensure the Board Understands the Critical Role of Resilience in Navigating Disruption

‍

Supply chain resilience initiatives often stall due to insufficient investment, mainly because boards lack a clear understanding of supply chain risks and their severity.

This led to a crucial question in the 2024 McKinsey Global Supply Chain Leader Survey: How well does your board understand supply chain risk and resilience?
‍

What we found, I must say, was quite shocking. Participants said that this year, only 36% of their board really understands what supply chain risk and resilience is about.
‍
‍Dr Knut Alicke
‍Partner at McKinsey and Company
‍

‍

‍

So What Can We Do?

‍

While Dr. Alicke's book Sourced to Sold provides comprehensive insights on this topic, supply chain leaders can follow two key pieces of advice:

• eliminate technical jargon that creates barriers between you and your audience,
• use storytelling to convey the importance of end-to-end operations and risk resilience.

The responsibility for better communication lies with both parties—we can't simply blame the board for not understanding. Since supply chain professionals often rely heavily on technical language, leaders must focus on clearer communication.

‍

‍

Hands-On Example: Car Insurance‍

‍

To illustrate his advice, Dr. Alicke uses the example of car insurance. We don't typically cut costs by eliminating car insurance because we clearly understand that we need it. We want protection in case of an accident.

‍

We need to be much better at telling this story to the board: "Look, this is the revenue at risk. This is the EBITDA at risk. These are the actions we need to take to mitigate that risk." For that, it's crucial to evaluate different scenarios. What happens if tariffs increase? What happens if there's a flood? We need to understand the breakpoints in the supply chain, develop mitigation levers, and evaluate their implications on revenue and EBIT.

Dr Knut Alicke
‍Partner at McKinsey and Company
‍

‍

Regarding supply chain leaders' tendency to use technical language that few understand, Dr. Alicke advises focusing on the outcomes when presenting to the board:

‍

Instead of saying, "We can increase our OTIF from 79% to 81%," let's say, "With these measures, we can improve EBITDA by a specific percentage.
‍
Dr Knut Alicke
‍Partner at McKinsey and Company
‍

‍

‍

Q&A

‍

Everybody's Talking About Generative AI: What Do You See as Its Potential Impact on Supply Chain Management?
‍

‍

Henrique Valer:

At a high level, AI and generative AI are the key technologies enabling Pelico to deliver the features we've discussed. This represents a true revolution, similar to the dot-com boom of the early 2000s or the smartphone revolution after 2008. GenAI will transform how we do everything.

‍

My advice is straightforward: if you're not familiar with it, get familiar. If you haven't started a GenAI initiative in your supply chain or planning, start now.

‍

Let me share a recent example. Last week, we tested some initial functionalities with a customer. Consider a simple question like "How many orders are related to this specific part?" In a traditional IT setup, this would require technical expertise—someone writing SQL or Python queries, running them, and visualizing the data. The process could take hours.

‍

With GenAI, anyone who can speak English—or any language—can ask the question and receive an immediate answer.

‍

However, an LLM alone isn't the complete solution—you need a comprehensive system to coordinate inputs, verify directives, and ensure accuracy. This is where Pelico bridges the gap between generative AI tools and the precise answers needed for daily operations.

‍

The impact is transformative. While traditionally only 2-3% of an organization can work with IT-heavy systems, these new tools enable 95-97% of the organization to access and use data. That's truly game-changing.

‍

Dr Knut Alicke:

‍

I fully agree with Henrique—GenAI will change how we operate supply chains, much like the invention of the container many years ago, which revolutionized global supply chains. GenAI is another positive disruption.

‍

However, the challenge is that GPT was released just two years ago, and now people are already saying, “We don’t see the business cases yet.” It’s not overhyped like blockchain was; it just needs time.

‍

Imagine GenAI as a new team member—fresh out of university, with no business background but access to all systems and data. You’d ask this team member, “Can you analyze customers with availability issues over the last six months and sort them by revenue?” GenAI can do that, but you still need to provide the business context.

‍

In three to five years, I expect GenAI to be widely adopted across planning functions. In the meantime, experiment with it. Try it out and learn. Tools like ChatGPT now even perform reasoning, not just pattern recognition.

‍

‍

What Supply Chain Risk Do You Find Most Frequently Underestimated, and What Causes Organizations to Overlook It?

‍

‍

Vera Trautwein:

‍

It's not the high-profile Black Swan events we often discuss that pose the greatest threat. The most damaging risks are those that develop gradually, almost imperceptibly.

‍

Consider the truck driver shortage—it didn't emerge suddenly but intensified over time until it affected the entire industry. The same pattern occurs when a supplier's delivery times slowly lengthen or product quality deteriorates incrementally. These small changes don't trigger immediate action because their individual impact seems minor.

‍

These creeping risks are particularly dangerous precisely because organizations tend to overlook them until they reach a critical point—and by then, it's too late to prevent the damage.

‍

‍

How to Get Customers to Share Sensitive Data, Especially Regarding AI and General Learning Models?

‍

‍

‍

Dr Knut Alicke:

‍

For starters, it's crucial to maintain a local version of these tools. Never upload your data to open platforms like ChatGPT—it's far too risky. Instead, host these tools in your local environment to ensure confidential data stays within your organization. This way, you can still benefit from trained large language models while maintaining security.

‍

Regarding data sharing with suppliers, which is a challenge that we've faced for 25 years at McKinsey and Company… While sharing fully confidential data isn't feasible, there are many opportunities for productive information exchange.

‍

Take your S&OP process, for instance. Why not integrate your suppliers? Since you need their commitment anyway, consider sharing volume forecasts or production plans. You can exchange valuable information while protecting sensitive data.

‍

The key is fostering collaborative relationships that benefit everyone.

‍

I've found success by first eliminating obstacles, particularly non-compete clauses with tiered suppliers. Then, establish a clear win-win narrative. While we seek data from suppliers, we must articulate what value we offer in return. This clear value proposition is essential for launching successful collaborations.

‍

Henrique Valer:

‍

I completely agree with what's been said. Let me share some real-world examples of data security concerns.

‍

We've witnessed serious incidents. For example, one of the earliest major IP breaches occurred when Samsung employees uploaded proprietary code to the public version of GPT, resulting in a data leak.

‍

Security challenges remain, especially regarding regulatory compliance. Take GDPR's "right to be forgotten" requirement—how do we implement this with AI systems that learn over time? Tracking data storage, learning processes, and timing becomes complex. We still need solutions for fully auditing these non-deterministic systems.

‍

For now, exercise extreme caution with data sharing practices. While secure versions exist to handle data safely, we must remain vigilant. Fortunately, we're making progress toward better solutions.

‍

The bottom line? Focus on locally hosted solutions and develop win-win scenarios for data sharing.

‍