MHI Cast

Supply chain leaders must distinguish actionable signals from data noise using systemic KPIs and AI context.

18 min Listen to the original Episode page
Supply Chain ManagementArtificial IntelligenceEconomic IndicatorsLabor MarketKPIsAutomationRisk Management

Summary

Hosts Michael Cavanaugh, Greg Barone, and Jason Schenker discuss how supply chain leaders can navigate data overload by focusing on critical metrics and leveraging AI for pattern recognition. Schenker emphasizes monitoring the MHI Business Activity Index and ISM manufacturing index while warning against suboptimizing individual KPIs in favor of a systemic view. Barone highlights the tight US labor market as a driver for automation in manufacturing and warehousing. Cavanaugh explains the importance of the 'perfect order' metric as a multiplicative aggregation of granular performance indicators. The speakers agree that AI must be guided by Standard Operating Procedures to provide context to frontline workers.

Key takeaways

  1. 1
    it also needs to add some sort of value to. service, economics, or risk mitigation. The difference between a one-time event or an interesting outlier and a signal is really the perseverance of that trend. Single point can be interesting and is worth paying attention to, but multiple points make a line, and if that line is pointed in a direction that you don't like, then that's something to take a look at.

    Jason Schenker · 7:13

Episode sections

  1. 0:04 – 2:25

    Introduction and Data Overload

    The host introduces the MHI Cast and the challenge of identifying actionable signals amidst data bombardment. Michael Cavanaugh emphasizes focusing on critical data elements and using human expertise to interpret AI outputs.

  2. 2:25 – 3:24

    Frontline Noise and AI Context

    Greg Barone notes frontline workers check 30-40 reports daily. He explains that AI systems must use Standard Operating Procedures (SOPs) to provide context and guide workers to high-value tasks.

  3. 3:24 – 4:56

    Key Economic Indicators

    Jason Schenker recommends monitoring the MHI Business Activity Index, ISM manufacturing index, and DC policy. He notes high interest rates and oil prices, with ISM weakness as a warning sign.

  4. 4:56 – 6:53

    KPI Selection and Systems Thinking

    Michael Cavanaugh advises focusing on output metrics and their leading indicators. Jason Schenker warns against suboptimizing individual KPIs, advocating for a systemic view that accounts for volatility.

  5. 6:53 – 9:19

    Signals vs. Noise and Geopolitics

    Schenker defines signals as repetitive trends adding value. Barone discusses long-term trends in AI productivity and geopolitical tensions between US allies and Cold War adversaries.

  6. 9:19 – 10:28

    Labor Market and Automation

    Barone highlights a tight US labor market with low unemployment despite slowed payroll growth. He argues this creates opportunities for automation in manufacturing and warehousing.

  7. 10:28 – 14:23

    AI for Pattern Recognition and Process Adherence

    Cavanaugh describes AI's role in flagging patterns. Schenker stresses the need for closed-loop AI learning and criticizes poor process adherence and lack of frontline leadership presence.

  8. 14:23 – 17:32

    Perfect Order Metric and Conclusion

    Cavanaugh explains the 'perfect order' metric as a multiplicative aggregation of granular metrics, where 95% performance across five metrics yields ~80% perfect orders. The episode concludes with a focus on critical output metrics.

Claims worth verifying

Prestige Economics has produced the MHI Business Activity Index for over a decade.

Jason Schenker ·3:24

How to verify: Check Prestige Economics website or MHI publications for the index's launch date and history. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

The MHI Business Activity Index provides the best near-term read on dynamics in terms of new orders, shipments, unfilled orders, inventories, and future new orders.

Unidentified speaker (UNKNOWN) ·3:38

How to verify: Compare MHI Business Activity Index release timing and content against other industry reports like the Federal Reserve's Beige Book or Census Bureau data. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

The MHI Business Activity Index is usually released just a few days after the end of each month.

Unidentified speaker (UNKNOWN) ·3:56

How to verify: Review the publication schedule of the MHI Business Activity Index on the MHI or Prestige Economics website. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Primary source

Interest rates are high and oil prices are high.

Jason Schenker ·4:16

How to verify: Check Federal Reserve interest rate announcements and Bloomberg/Reuters oil price data for the verification period. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Consumer confidence is a bit weaker.

Jason Schenker ·4:16

How to verify: Check the Conference Board Consumer Confidence Index or University of Michigan Consumer Sentiment Index for the verification period. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

The US unemployment rate is relatively low.

Unidentified speaker (speaker_2) ·9:39

How to verify: Check the US Bureau of Labor Statistics (BLS) Current Employment Statistics (CES) report for the verification period. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Payroll growth has slowed.

Unidentified speaker (speaker_2) ·9:39

How to verify: Check the US Bureau of Labor Statistics (BLS) Nonfarm Payrolls data for the verification period. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

If five metrics perform at 95%, the perfect order aggregation is going to be something around 80%.

Michael Cavanaugh ·14:41

How to verify: Calculate the mathematical result of 0.95^5 to verify the multiplicative effect on the perfect order metric. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

The MHI Business Activity Index is produced by Prestige Economics.

Jason Schenker ·3:24

How to verify: Check the copyright or authorship information on the MHI Business Activity Index report or Prestige Economics website. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Do your own research

  • Verify the current value and trend of the ISM Manufacturing Index via the Institute for Supply Management website.
  • Check the latest MHI Business Activity Index data from Prestige Economics to assess near-term industry dynamics.
  • Review US Bureau of Labor Statistics data for recent payroll growth and unemployment rates to validate labor market tightness.
  • Analyze company-specific SOPs to determine if AI implementations are context-aware or merely reporting raw data.
  • Assess the 'perfect order' calculation methodology for key suppliers or competitors to understand true operational efficiency.
  • Monitor US federal policy announcements for potential exogenous shocks to supply chain operations.
  • Evaluate the correlation between oil price fluctuations and manufacturing index weakness in historical data.
Full transcriptexpand
0:04
Unidentified speaker (speaker_0): Welcome to the MHI Cast, the show where we talk to the industry's best minds to uncover their supply chain stories. We explore real-world case studies and get unique perspectives on key trends and emerging technologies from every corner of the material handling, logistics, and supply chain space.
0:30
Unidentified speaker (speaker_0): Economic instability, geopolitical uncertainty, new technologies, shifting customer demands. Supply chain leaders have more on their plate than ever. But with so many competing indicators, how do you know which ones to focus
0:45
Unidentified speaker (UNKNOWN): on? The
0:46
Unidentified speaker (speaker_0): challenge isn't necessarily getting more data, it's identifying the signals that matter. understanding what they're telling you, and knowing when and how to act on them. Joining us to explore these challenges are Michael Kavanaugh, CEO of KPI Solutions, Greg Barone, Vice President of Strategic Accounts at Smart Access,
1:05
Unidentified speaker (speaker_0): Jason Shanker, President of Prestige Economics, and Will Sparks, Vice President of Supply Chain Solutions at Kenco Group.
1:14
Unidentified speaker (UNKNOWN): Okay,
1:14
Unidentified speaker (speaker_0): Michael. let's start with the problem itself. When leaders are surrounded by so much data, how do they identify what's actually actionable?
1:23
Michael Cavanaugh: Yeah, I think if anything, it's almost bombardment with data, right? There are a few things. One is trying to figure out the critical few, right? There are usually a handful of data elements that really help you make a decision about whatever operational improvement. you need to make. And I think being able to focus on what are the data that's
1:43
Michael Cavanaugh: actually going to drive a change or a decision, start there. And the other thing is the people who have expertise and the ability to apply critical thinking to the data that they're seeing. You can throw a lot of data at AI and get some really interesting ideas back. but
2:03
Michael Cavanaugh: it takes that experience and that critical thinking together that really helps you make the right decision and guide the decision in the direction for success.
2:13
Unidentified speaker (UNKNOWN): On
2:13
Unidentified speaker (speaker_0): the warehouse floor, leaders navigate dozens of reports and dashboards, often while making decisions in real time. Greg, what is does that noise and information overload look like for frontline workers?
2:25
Unidentified speaker (UNKNOWN): You're
2:26
Unidentified speaker (speaker_2): right. There is a lot of noise, and there's a lot of noise before AI. We're hearing from frontline leaders that they're having to check 30 to 40 different reports throughout the day. And AI can generate even more information and more signals. And so it's important for the AI system to that have
2:46
Unidentified speaker (speaker_2): context of what's important and what's not important. We in our
2:51
Unidentified speaker (speaker_2): platform allow for SOPs to be loaded in so that you can have that context of what's important, and that the system then can guide that frontline leader where they should be spending their time.
3:04
Unidentified speaker (speaker_0): So the goal isn't to monitor everything, it's to narrow the focus to a smaller set of indicators that can reveal where the business is heading. At the industry level, that means understanding which economic signals can warn of changing conditions early. Jason, what should supply chain leaders be watching most closely?
3:24
Jason Schenker: Supply chain leaders should watch the monthly MHI Business Activity Index by Prestige Economics very closely. This is an index my firm has produced for over a decade for this industry, and it provides
3:38
Unidentified speaker (UNKNOWN): the industry.
3:39
Jason Schenker: the best near term read on dynamics in terms of new orders, shipments, unfilled orders, inventories, and future new orders. That is the best data available to see if there is a shift in this industry. Shift higher, shift lower.
3:56
Unidentified speaker (UNKNOWN): It
3:56
Jason Schenker: will show up in that data very quickly, which is usually released just a few days after the end of each month. That's going to be the quickest data specifically for this industry. I would also encourage people to pay very close attention to the ISM manufacturing index. It has been very strong in recent months, but
4:16
Jason Schenker: interest rates are high and oil prices are high, and consumer confidence is a bit weaker. If the ISM weakens, that would be a warning sign. And the third thing that people should watch closely is policy out of DC. It is an exogenous risk outside of their control that, as we've seen in recent
4:36
Jason Schenker: years, can very quickly change the path and strategy and operations of a business almost overnight.
4:45
Unidentified speaker (speaker_0): The right indicators can depend on what the organization is trying to achieve. Michael how should leaders determine which KPIs deserve the most attention for their specific situation?
4:56
Michael Cavanaugh: The relevance of a KPI comes down to what action am I going to take because of that KPI. If you think about it, I'm an old Six Sigma guy. So you have your inputs and you have your outputs. And really understanding the transfer function, what are the inputs that drive a particular output? To me, the red flags should be looking at what is the output? Am I going to get
5:16
Michael Cavanaugh: the output that I expect based upon the inputs? And getting too focused on some of the input metrics can mean that you lose sight of what the big why is. What is it you're really trying to accomplish with your operation? And so I would say you've got to have the right balance of here's my output metric that
5:36
Michael Cavanaugh: I'm really focused on, and here are the handful of leading indicators for that particular output metric. And if you understand your operation well enough to know what your leading indicators are that are affecting your priority output metric, then those are the metrics you really need to hang your hat on.
5:54
Unidentified speaker (speaker_0): Of course, individual KPIs don't always tell the full story. An operation might be performing well in one area while another metric reveals. a problem. Will, I'd love to hear from you. Why is it important to look at these indicators together rather than in isolation?
6:11
Unidentified speaker (UNKNOWN): I
6:11
Jason Schenker: think it's important that supply chain leaders look at their KPIs as a system. I think suboptimizing KPIs can present potential inefficiencies, and you could risk the entire system's effectiveness by trying to suboptimize one metric. So I think the supply chain metrics that we have developed over
6:31
Jason Schenker: decades are good, but I don't think that they're complete. I think a lot of times we focus on averages, and we need to fully understand the volatility and variability behind those averages.
6:43
Unidentified speaker (speaker_0): So how should leaders distinguish between an isolated event and a meaningful trend? What separates a signal worth? acting on from everyday noise?
6:53
Jason Schenker: There's a lot of noise in supply chain right now, especially with the growth of AI. And I think it's important that we focus on those elements that become signals versus noise. And I think something becomes a signal when it's repetitive and it's not an outlier or a one-time event. And
7:13
Jason Schenker: it also needs to add some sort of value to. service, economics, or risk mitigation. The difference between a one-time event or an interesting outlier and a signal is really the perseverance of that trend. Single point can be interesting and is worth paying attention to, but
7:33
Jason Schenker: multiple points make a line, and if that line is pointed in a direction that you don't like, then that's something to take a look at.
7:40
Unidentified speaker (speaker_0): That distinction matters most when signals involve economic uncertainty, geopolitical events, or emerging technologies. Jason, how can leaders avoid getting distracted by the latest headlines and focus on the developments that should actually change their strategy?
7:57
Unidentified speaker (speaker_2): To separate signals from noise,
8:01
Unidentified speaker (speaker_2): I often think about something I heard when I worked on the trading desk of an investment bank.
8:07
Unidentified speaker (UNKNOWN): And
8:07
Unidentified speaker (speaker_2): the saying was: the trend is your friend.
8:11
Unidentified speaker (UNKNOWN): That
8:12
Unidentified speaker (speaker_2): doesn't mean the trend is always friendly,
8:15
Unidentified speaker (speaker_2): but it's really important to watch those longer-term trend lines and not just the short-term headlines.
8:23
Unidentified speaker (UNKNOWN): When
8:23
Unidentified speaker (speaker_2): I think about AI, I think about
8:26
Unidentified speaker (UNKNOWN): a
8:27
Unidentified speaker (speaker_2): long-term trend of applying new technologies to boost productivity. Something that's been happening since we were cavemen.
8:37
Unidentified speaker (speaker_2): When I think about geopolitics, I think about rising tensions between the U.S. and its allies on one side and some of our former Cold War I adversaries on the other.
8:52
Unidentified speaker (speaker_2): And the tensions we've seen rising in recent years in terms of trade, supply chains, and geopolitical integration among those allies on both sides.
9:05
Unidentified speaker (UNKNOWN): And
9:05
Unidentified speaker (speaker_2): I think about that trend and what it would take to get that trend to reverse or even to get it to slow down.
9:13
Unidentified speaker (speaker_0): Then how should supply chain leaders think about risk when deciding where to focus their attention?
9:19
Unidentified speaker (speaker_2): Not all risks are bad. There are risks to the US outlook or the global outlook or metals prices or energy prices, but many of those risks can often contain embedded opportunities, automation to augment.
9:39
Unidentified speaker (speaker_2): a very strapped labor market in the US. Even though we hear that payroll growth has slowed, which it certainly has, the unemployment rate is relatively low. The labor market is tight. And demands in manufacturing, material handling, warehousing across the supply chain and utilities are going gonna to continue to rise.
9:58
Unidentified speaker (speaker_2): Automation in these sectors is going to be critical. So you see the risks to embrace are focusing on these areas of opportunity that go beyond the headline.
10:11
Unidentified speaker (speaker_0): The same principle applies inside the warehouse, where recognizing an emerging problem early can create an opportunity to respond before it affects performance. Michael, what does that look like when you're making decisions about orders, inventory? and fulfillment?
10:28
Michael Cavanaugh: I think it varies definitely by level within the operations as well, right? You definitely have some people who are functioning on the floor level, for example, looking at the journey of an order, and they're getting all kinds of data about is this order going to be ready for shipment? Do we have all the inventory necessary to complete that order? They're seeing all of
10:48
Michael Cavanaugh: that. They've got cues of orders in priority. order. They've got to make kind of fast decisions about whether or not something that's a priority order is going to be able to be filled. What if a new priority order comes in faster than that and you've got an inventory challenge? And I think today a lot of that decision making is happening at the manual
11:07
Michael Cavanaugh: level. And I think what the systems have the ability to do is take a lot of that data and rather than just throw the data at the operator, give them the flags to be looking for and actually bring up the fact that a new order needs their attention or it's going to be late. And rather than let them
11:27
Michael Cavanaugh: find out when it's hit a roadblock. And I think today, too often, orders will We'll hit a roadblock, and then someone has to step in and try to fix it. The goal would be try to. preempt a roadblock and give people the information that they need in order to address that.
11:44
Unidentified speaker (speaker_0): How can AI help leaders identify warning signs they might otherwise miss?
11:50
Michael Cavanaugh: One of the masterful things about AI is its ability to crunch through massive amounts of data and find correlations and similarities in patterns. And to me, that's something that we really need to leverage AI for more. So if it's seen this pattern over and over again and it
12:10
Michael Cavanaugh: sees that pattern shaping up again, it's a great red flag. Does it mean that you should let that make it the decision for you? No. But it's a great opportunity for AI to course through all of that data, look for those patterns, and raise that flag to the operator who. has the expertise to then step in and make
12:29
Michael Cavanaugh: a decision.
12:31
Unidentified speaker (speaker_0): That human judgment is essential. Technology might recognize that something is changing, but people still need to understand why and whether it warrants intervention. Greg, how can AI help frontline workers spot when a process starts to deviate from expectations?
12:48
Jason Schenker: AI, if it has the context of the facility, what should be performed within the facility, and it could leverage that information off of work instructions and SOPs, as well as job functions within the facilities. And the system will then be able to provide guidance as to if
13:08
Jason Schenker: a process is deviating and by how much it's deviating. There's also the ability
13:15
Jason Schenker: for human in the loop. So as the frontline leader goes out and engages someone and corrects it, they can take notes and information that the AI system will then learn from to determine if that's a valuable thing that frontline leaders should be spending time on.
13:32
Unidentified speaker (UNKNOWN): How
13:32
Unidentified speaker (speaker_0): do you make sure the signals being delivered actually lead to meaningful action?
13:37
Jason Schenker: The system's got to have gotta content. have context. And you don't want to provide too many signals because a signal without action is really a waste of resources. So you want to make sure that you provide the highest value signals for those frontline leaders to go out and engage. But just providing the signal is one thing, but making sure that the action actually
13:57
Jason Schenker: happens and what is the outcome of the action, that's all information that needs to get back into the AI. so the AI system can continue to learn from that. We feel that that should be a closed loop process, and that's really important from an AI perspective.
14:11
Unidentified speaker (speaker_0): Leaders need to stay focused on key objectives rather than chasing every metric. Will, is there one measure that can help bring those different aspects of supply chain performance together?
14:23
Unidentified speaker (UNKNOWN): Well, I think
14:24
Unidentified speaker (speaker_2): the ultimate metric in supply chain is this concept of a perfect
14:28
Unidentified speaker (UNKNOWN): order. And
14:29
Unidentified speaker (speaker_2): it's hard
14:29
Unidentified speaker (UNKNOWN): to
14:30
Unidentified speaker (speaker_2): reach
14:30
Unidentified speaker (UNKNOWN): that
14:30
Unidentified speaker (speaker_2): metric at 100% because it's
14:33
Unidentified speaker (UNKNOWN): the
14:33
Unidentified speaker (speaker_2): multiplication of multiple
14:36
Unidentified speaker (UNKNOWN): more
14:36
Unidentified speaker (speaker_2): granular
14:37
Michael Cavanaugh: metrics. And so even if you've got five
14:40
Unidentified speaker (UNKNOWN): metrics that are
14:41
Michael Cavanaugh: performing
14:41
Unidentified speaker (UNKNOWN): at 95%,
14:43
Michael Cavanaugh: your perfect
14:44
Unidentified speaker (UNKNOWN): order in aggregation is going to be something around
14:46
Michael Cavanaugh: 80%.
14:47
Unidentified speaker (speaker_0): Greg, where do you see organizations missing important? opportunities to improve performance?
14:53
Jason Schenker: Process adherence. And we see that frequently within facilities. Some organizations will perform process observations. Some will only perform them on low performers. Some won't even perform them at all. And we see that constantly throughout organizations and paying more attention
15:13
Jason Schenker: to the process adherence.
15:15
Jason Schenker: we have seen can improve the performance of the facility significantly. I think that they believe that the actions are happening. I think about it from a sports perspective. So you're working in these facilities and you have industrial athletes who are out and performing the work that needs to get done. And if you're running a
15:35
Jason Schenker: professional team, you would never not
15:37
Unidentified speaker (UNKNOWN): go
15:38
Jason Schenker: to practice. You wouldn't spend a practice time or a game time back in the office. And that's happening frequently. You need to spend that time out on the floor. And so a lot of time we hear from customers that the frontline leaders are not on the floor. And when that happens, you get deviations in the process. And we see it frequently.
15:57
Unidentified speaker (speaker_0): As we wrap up this conversation, Michael, for leaders looking to simplify the information. they're monitoring, what's the most important thing to get right?
16:06
Michael Cavanaugh: I do think it comes down to understanding your critical output metric. Maybe it's three, maybe it's five, but what are really the critical metrics that are driving your operation? And then what are the leading indicators? And if you can really kind of pace paste that line between those two for your inputs, you've got a great
16:27
Michael Cavanaugh: way of saying, if I want that outcome, here are the handful of levers I can pull. So if one of them becomes a roadblock or a problem, I can go to one of the others to try to change that output. To me, that's sort of the magic of it. And I think AI has helped us to see those patterns much more
16:46
Michael Cavanaugh: clearly.
16:49
Unidentified speaker (speaker_0): Finding the signal in the noise is isn't just about predicting what happens next. It's about making better decisions with the information you already have. And that means knowing which indicators directly impact your operation, recognizing when a trend signals a meaningful change, and understanding what actions you can take. Thanks to Michael, Greg,
17:09
Unidentified speaker (speaker_0): Jason, and Will for their insights, and thanks to you, our audience, for listening to this MHI cast. At MHI, we never stop exploring new opportunities to help you take your manufacturing and supply chain operations to the next level of success. Thanks for making us part of your professional development journey.
This page was produced automatically: the audio was transcribed with nvidia/parakeet-tdt-0.6b-v3+sortformer-v2.1 and interpreted by Qwen3.6-35B-A3B-UD-IQ3_S.gguf. Transcription and interpretation can contain errors; always confirm against the original episode. Nothing here is investment advice.
All insights