Sean Lefever @SeanLLefever

A common assumption in forecasting: If we just had better data, the forecast would improve. Better data absolutely helps. But in complex environments like automotive, it’s usually not the root issue. What we tend to see instead is: • Data is fragmented across systems • Assumptions aren’t aligned across teams • Changes happen faster than models can adapt So even with more data, the outcome doesn’t always change in a meaningful way. The challenge is less about the amount of data and more about how it’s used, aligned, and updated over time. Improving forecasting isn’t just a data problem. It’s a coordination and adaptability problem. Would you agree?

One of the biggest shifts in automotive right now isn’t just demand volatility. It’s how quickly planning assumptions become outdated. In many environments, plans are still built around relatively fixed cycles: • Monthly updates • Quarterly resets • Annual planning processes But the operating environment doesn’t move on those timelines anymore. What this creates is a constant gap between: • What teams planned for • What they’re actually experiencing And that gap tends to show up downstream—in missed targets, reactive decisions, and increased pressure across teams. The organizations that seem to navigate this best aren’t necessarily the ones with more data. They’re the ones that are better at: • Adjusting assumptions quickly • Aligning teams around changing signals • Treating planning as an ongoing process, not a fixed event Curious how others are seeing this play out.

I hear you and that is a real problem in a lot of industries. But this is a different situation. This is not a vendor cutting headcount to boost their own margins. This is a manufacturer whose team shrank from 33 to 12 over several years through attrition, retirements, and budget pressure. The work did not shrink with them. Same number of programs. Same number of customers. Same number of price negotiations and engineering changes. The 12 people left are doing the work of 33 because the platform absorbed the manual effort that used to require bodies. Data entry that took a team of seven is now maintained by one person. Portfolio impact analysis that used to take weeks of spreadsheet work happens in seconds. That is not doing less with less. That is doing the same work without burning out the people who are left. The real question for most manufacturers is: when your team shrinks (and it will, it is happening everywhere), does the work get automated or does it just pile up on fewer desks?

One of our customers used to need 33 people to manage their commercial lifecycle across programs. Today they do it with 12. One person now maintains what seven people used to contribute to. Not because they hired better. Because the system replaced the manual work that was eating their team alive. The same person can select every EV program in their portfolio, zero out a volume scenario, and see a $250 million impact in seconds. That is not a reporting tool. That is a decision engine. And when their global headquarters saw it, the reaction was: “Nobody else is doing this. Not even us.” The best compliment a platform can get is when someone says “without this system, we cannot survive.” That is what we heard this week. What would your team do with 60% fewer people on manual commercial work and 100% more visibility into program impact?

@Kylel999 Nice try

Your demand planner says 10,000 units. Your sales team says 14,000. Your customer EDI says 8,500. Everyone is working off a different number. And when the actual lands at 9,200, the cascade starts: -> $3M in premium freight to cover the gap -> $4M in excess inventory from the overforecast -> Overtime costs nobody budgeted for The problem is not that forecasts are wrong. They will always be wrong. The problem is nobody is measuring how wrong and what that variance costs on each program. What is your forecast-to-actual variance? And what does it cost you?

This is exactly what's playing out in automotive supply chains right now. Lear cut 15,000 jobs last year. Bosch: 13,000. ZF: 14,000. Ford just announced 15,000 more. That's not headline panic...that's capital math catching up. But here's what nobody's tracking: each person who walks out takes institutional knowledge with them. Which programs are profitable. Which suppliers deliver. Which cost assumptions always break post-award. 70% of that knowledge is undocumented. It lives in someone's head and a spreadsheet they built in 2011. So the capital math forces the layoff. But the layoff creates a second problem nobody budgeted for which is the next person who touches that program starts from zero. And "starting from zero" on a $50M program isn't a learning curve. It's margin leakage from day one. The workforce doesn't just move slowly. It takes the margin intelligence with it when it goes.

The question nobody asks: what happens to the cost that CAN'T be passed through to consumers? Suppliers absorb it. Industry EBIT margins are already at 4.7%. European suppliers: 3.6%. There's no buffer left. Analysts are predicting a wave of supplier bankruptcies in 2026. The ones who survive see margin impact faster...not just cut costs faster.

The just-in-time point is critical. Buffer inventories have been thinned by years of lean philosophy. They're not sized for 14 extra days of transit time. But the bigger question for suppliers: which PROGRAMS just went negative? Oil spike + shipping reroute + tariff stack = three cost shocks on the same programs simultaneously. Most manufacturers won't know the answer for weeks. That's the real risk.

Four weeks is the bet. But the semiconductor shortage was a “12-week problem” that lasted two years. Red Sea disruptions were “temporary”and carriers are still rerouting. For manufacturers, the question isn’t how long Hormuz lasts. It’s how fast you can see which programs just went negative. A $2B supplier just had the cost basis shift on every single one of its programs. At industry margins of 4.7%, there’s no buffer to wait and see. The suppliers who quantify program-level impact in days/hours will recover costs.

What’s missing from the container shipping conversation: automotive parts. Rerouting via Cape of Good Hope adds 10-14 days per shipment. Assembly plants in the US, Mexico, and Germany will feel delayed Asian component shipments within 2-3 weeks. And the semiconductor shortage started as a “12-week problem.” It lasted two years.

The biggest risk isn’t higher costs…it’s the delay in knowing WHICH programs just went underwater. A $2B supplier with 200 active programs just had the cost basis shift on every one. Some were already marginal. Most won’t know which ones for weeks. Speed of response is the real competitive advantage here.

Oil spiked 10%. Shipping costs doubled. Supertanker rates hit $423K/day which is an all-time record. Somewhere inside every auto supplier right now, someone is trying to answer: “Which of our 200 programs just went underwater?” Most won’t know for weeks. By then, the margin is gone.

A $2B Tier 1 supplier absorbed 200+ engineering changes last year between award and SOP. Average cost per change: $25K. Total: $5M. Amount recovered from the customer: $0. Not because recovery was not possible. Because nobody was tracking the cost per change in a way that made the recovery conversation defensible. How many design changes did your team absorb last year? And does anyone know the number?

That’s actually more P&L visibility than most people in your role get — and the fact that you’re tracking COPQ and OT per program is significant. Here’s the question I’d ask the finance team: are they pulling your COPQ and OT data into the margin walk on those programs? Because if you’re tracking $200K in quality costs and $150K in unplanned overtime on a specific program, and finance is still showing the original awarded margin in their forecast — those two numbers aren’t talking to each other. That’s where margin erosion hides. The awarded margin assumed a certain quality yield and a certain labor plan. Your actuals tell a different story. If nobody is reconciling the two, the program looks profitable in the ERP and unprofitable on the floor. You’re closer to the margin problem than you think.

One of our clients — a $4B global manufacturer — was bleeding margin in 3 places: → $12M in premium freight from forecast misses → $14.4M in excess inventory carrying costs → $20M in margin erosion they couldn’t see Total: $46M+ per year. They didn’t need more data. They needed margin certainty. That’s what we gave them. If you’re running a manufacturing business above $500M, these numbers probably look familiar. What’s your biggest margin leak? Drop it below — I bet the pattern is the same.

That makes sense and honestly, the people doing what you do see the margin problem more clearly than most P&L owners do. They just see it from the operational side. When you’re training engineering talent and improving efficiency, you’re solving the same problem we are but just from a different angle. The gap we see is that the efficiency gains your team creates often don’t flow through to the financials because nobody is connecting the operational improvement to the program-level margin impact. Curious…when your team improves OEE or reduces rework, does the finance side ever quantify what that saved on a specific program? Or does it just show up as a general improvement in the quarterly numbers?

You’re 100% right — and that’s exactly the tension. Finance sees $14M in carrying costs and wants to cut inventory. Supply chain knows that if they do, one missed shipment cascades into premium freight, missed delivery windows, and penalty charges that dwarf the carrying cost savings. The problem isn’t that either side is wrong. The problem is they’re making those decisions with different numbers. Finance is looking at a quarterly snapshot. Supply chain is looking at supplier reliability history. Nobody is looking at the actual program-level margin impact of both scenarios side by side. That’s the gap — not too much inventory or too little, but no shared view of what the tradeoff actually costs on a specific program. What’s your experience — does your team have a way to quantify that tradeoff, or is it still a gut call?

China setting the OEE benchmark is real….and it creates an interesting problem for the rest of the portfolio. When one region is running 85%+ OEE and another is at 65%, the margin walk on the same program looks completely different plant to plant. Most companies we work with can tell you their OEE by line, but they can’t tell you how that variance is actually flowing through to realized margin on a specific program. That’s where the visibility gap lives. Are you seeing that — where the operational metrics look fine in one region but the financials don’t match?

Quality and efficiency…that’s what we hear most often. The interesting thing is when we dig in, those two usually trace back to the same root cause: the forecast-to-actual gap. When volume assumptions are off, you’re either expediting (killing efficiency) or building to the wrong spec window (killing quality). Where are you seeing it hit hardest? On the floor or in the planning process?