Seven Insights From Supply Chain Experts on Faster Decision-Making

Picture a tough question coming in at 8 AM on a Monday: “what does the new tariff do to our costs, and what should we do?” The old way involves weeks or months of tedious data gathering and scenario analysis. Today, it’s possible to ask a question at 8 AM and get a trusted answer by lunchtime, using any familiar AI interface.  

We surveyed 144 senior supply chain professionals with Supply Chain Media to get a clear picture of how global organizations are making supply chain decisions. Only 18% can evaluate a major decision in days or hours. The other 82% need a week or more, and a quarter need longer than six weeks.

Why are some companies ready to act as soon as a question arrives, and some are still waiting for an answer weeks later?

We took those results and more from our full Supply Chain Design & Resilience Survey report to a live panel with:

  • ‍Rob Carlisle, Associate Partner at Argon & Co.
  • ‍Andrea Paciaroni, Network Strategy Practice Lead at Accenture ‍
  • Christophe Frère, VP, EMEA at Optilogic ‍
  • Martijn Lofvers, Founder and Chief Trendwatcher at Supply Chain Media

Here are the seven top takeaways from our expert panel.

1. Geopolitics still tops the list of supply chain disruptions

Geopolitics may have slipped down the CEO's earnings-call agenda, but the panel was clear it has not gone away. It has become, as Andrea Paciaroni of Accenture put it, "the new normal."  

Rob Carlisle of Argon & Co concurs. “Geopolitics is 100% in the supply chain directors’ and the operations directors’ short list of concerns because they deal every day with tariffs, lead times and volatile fuel costs.” The survey agrees: geopolitical instability (68%) and tariffs and trade restrictions (64%) topped the list of disruptions that hit organizations hardest.

The effects also travel further than most plans assume. The war in Ukraine exposed ties to wheat, glass and other primary materials, and, as Rob said, "I don't think that's necessarily something that supply chain leaders realized at the time—how exposed they were in certain areas to conflicts around the world." Even conflicts that seem far from the US now show up in gas prices and shipping, which is why sensitivity analysis keeps coming up with clients.

Main point: when disruption is a permanent input, the question is no longer whether it will hit. It is how quickly you can respond to the next one.

2. The inventory buffer is a good first move and a bad last one

The most common response to disruption in the survey was adding inventory (28%). Only 21% said they had changed their supply chain continuously over the past two years, and only 16% put more effort into the supply chain they will need than the one they run today.

The panel did not dismiss the inventory buffer. As Rob put it, "It protects your service, protects customers, and buys you a bit of time." The risk is what comes next: "it is really easy for that inventory buffer to become permanent." And it does not settle the bigger question, because "inventory buys you time, but it doesn't answer the long-term question: Is my network in the right place to allow me to respond to the challenges that I'm facing?"  

For Andrea, the gap between 21% and 16% points to a way forward: "I need to start now redesigning the supply chain that I need for tomorrow to enable more flexibility in the future." Christophe Frère of Optilogic put the stakes simply: "If you keep on doing the same thing you did yesterday, you're not improving." Host Martijn Lofvers added that supply chain should be a facilitator of growth, not just a team that fights fires.

"It protects your service, protects customers, and buys you a bit of time...[but] it is really easy for that inventory buffer to become permanent." Rob Carlisle, Associate Partner at Argon & Co.

Main point: the buffer buys you time. Use that time to design the supply chain that won’t require that extra inventory.

3. Three things constrain speed to action

Even when leaders agree they need to move faster, three bottlenecks stand in the way, as Rob laid out:

  1. Data. Finding it "is still hard," especially in companies that have grown by acquisition. Large multinationals routinely run 20 or 30 ERP systems.
  2. The model solve. This roadblock is fast disappearing: "in the last 2 years that that time is just collapsing. Today you can run a complex model in minutes; hours at most."
  3. People. Once the first two speed up, the limit becomes the human ability to digest the outputs. "You can almost have paralysis by insight.” The key is to use AI to automate the tedious portions and recommend solutions so that humans need only review and make the call.

The fastest-moving businesses, Rob observed, build this capability in-house, because their teams live with the data and the questions every day. As Andrea noted, "We often see companies dragging on even critical decisions for months, when the aim should be making key decisions in a week or less." AI can help on both sides of the decision: preparing the information and interpreting what the analysis produces.

"We often see companies dragging on even critical decisions for months, when the aim should be making key decisions in a week or less." - Andrea Paciaroni, Network Strategy Practice Lead at Accenture

Main point: as tested answers are now ready in hours, the constraint moves from the model to the people. Leverage AI to make teams faster, while preserving the human ownership of critical decisions.

4. Leaders need a directional answer; then analysts can refine it

The decision has moved up the organization: 35% of leaders say executive leadership owns supply chain design decisions, more than any dedicated team. “The people making the call may not have supply chain expertise, so recommendations must be easy to interpret," said Andrea. “The executive wants to know simply, "Which scenario are we are in now, and which we should consider next?”  

The ability to answer tough supply chain questions is being democratized: "We are seeing new decision-making capabilities being built at all parts of the organization," Rob said, and the businesses that respond fastest bring leadership decisions with real confidence knowing robust evidence sits underneath.  

And asking is getting easier. Christophe described how “an executive used to send a question to the operations research team and wait weeks or months for a single answer. Today they can ask an AI assistant like Claude or ChatGPT or another tool they already use what a 30% price increase would do to supply chain costs and get several tested scenarios back right away.” He added that you still need a model with your own supply chain data and supply chain blueprint underneath. Without one, the AI answer isn’t grounded in your own data.

“An executive used to send a question to the operations research team and wait weeks or months for a single answer. Today they can ask an AI assistant like Claude or ChatGPT or another tool they already use what a 30% price increase would do to supply chain costs and get several tested scenarios back right away.” - Christophe Frère, VP, EMEA at Optilogic

Main point: executives get a directional answer fast. Modelers and planners refine and validate it. AI speeds up the entire decision-making process.

5. Good-enough data, used now, beats perfect data later

Data came up as the leading limitation: 70% cited data quality among their biggest modeling limitations, 52% cited a lack of integrated systems, and only 22% named budget.

Rob offered a useful nuance: "Surprisingly, we don't find data quality to be too bad. But actually finding the data itself can be difficult, especially across fragmented systems left over from acquisitions.” His advice is to avoid perfectionism: "This isn't a data vanity exercise. It doesn't need to be perfect, and it never will be." Sometimes it is better to be roughly right and move at speed than to wait to be precisely right and decide too late.

"This isn't a data vanity exercise. It doesn't need to be perfect, and it never will be." - Rob Carlisle, Associate Partner at Argon & Co.

Main point: start with the data you have, agree on the assumptions, and keep it refreshed instead of rebuilding it for every question.

6. From belief in AI to everyday use

Leaders believe in AI: 70% expect it to help most with scenario generation, and 60% with data cleansing and integration. But only 4% use AI extensively across multiple processes, and 16% use AI-driven modeling tools.

Closing that gap takes three things, according to Rob:

  • The data underneath is trusted. AI can only reason as well as the data it is fed. ‍
  • The models and baselines are calibrated to reality. "AI can hallucinate at times," so it needs to be grounded in reality. ‍
  • Human judgment stays at the forefront. "The human judgment is still essential, and teams should focus above all on the decision you are trying to make, instead of getting wrapped up in the volume of scenarios.”

Expectations are high, and Andrea expects usage to grow quickly. He sees three barriers: access (software that embeds AI, so teams do not need extra architecture is a plus), trust, and the learning curve.  

Main point: AI speeds up good decisions. It does not replace the data, the model, or the people.

7. Fast, confident decisions require a capability, not just a tool

The survey found that teams running design as a continuous, embedded practice are far more likely to evaluate a major scenario in under two weeks (79%) than project-based teams (30%). Among the most mature organizations, 41% run continuous scenario analysis, against 16% of the rest.

Mature organizations do not just own design software. As Andrea described it, they build an in-house capability, a center of excellence for decisions about network flows and inventory positioning. With the right tool, team and AI, the software stops being something you open once a year or once a quarter. Instead, "I have this digital tool that everybody in the organization knows is available for continuous decision evaluation."  

"I need to start now redesigning the supply chain that I need for tomorrow to enable more flexibility in the future." - Andrea Paciaroni, Network Strategy Practice Lead at Accenture

Main point: the teams that move fastest treat design as a continuous process instead of a project.

Three panelist recommendations you can start now

  1. Start with one decision. Focus on the decision you are trying to make, not how many scenarios you could run.
  2. Use the data you have. Identify the data needed to address your priority decision. Agree on the assumptions, run the scenarios, and sharpen the answer and add additional data as you go.
  3. Make it a habit. Refresh the picture of your supply chain with a digital twin instead of rebuilding it each time, so the next tough question starts from a tested baseline.

The survey behind this conversation, The Supply Chain You Have vs. The One You Need, was run by Supply Chain Media with Optilogic. Download the full report

‍

No items found.

Latest Updates from Optilogic

Latest Updates from Optilogic

No items found.
No items found.
No items found.
No items found.