Hardware Managers Are Killing Automotive Software

Welcome to Issue #96 of The German Autopreneur.

For over a century, cars were defined by their mechanics. Horsepower. How they felt on the road.

That's changing.

Cars are becoming computers on wheels. Software Defined Vehicles (SDVs). And legacy automakers struggle with this shift. Not because of technology. Because of their own structures.

• "We need 40,000 new software developers"

• "We're launching a new software unit"

You hear this from boardrooms constantly. The assumption: If we hire enough people, we'll crack it.

But here's what Michael Fait told me. He works with automakers worldwide on building SDVs:

"Writing code isn't the problem in automotive software."

If code isn't the problem, what is?

Traditional automakers are trying to build software products using hardware company structures. That's the core issue.

I recently spoke with Michael about this. Today, I'm sharing the key insights from our conversation. Why middle management blocks innovation without meaning to. Why a hardware manager saving 50 cents can cost $100 million. And why the solution isn't more developers. It's different leadership.

100 Computers Must Work in Perfect Sync

First, let's understand the scale.

A modern car isn't one computer. It's a network of roughly 100 small computers.

These are called ECUs (Electronic Control Units). They control everything. Brakes. Airbags. Radio. Even window motors. Functions kept getting added over the years.

Millions of lines of code run on these ECUs. And all 100 computers must communicate in real time. Without errors. Because mistakes can hurt people.

Think of it as a digital orchestra playing in perfect sync at highway speeds.

Michael puts it this way:

"If someone had told me: We're building 100 different software services, deploying them all to the internet, and they need to work together flawlessly in real time... I would have said: That sounds like a pretty stupid idea."

But that's exactly what happens in cars today.

You can't solve this complexity by hiring more developers. You solve it by reorganizing how companies work.

And that's exactly what blocks legacy automakers. Here are the three reasons.

1) The 50-Cent Dilemma (or: Why Hardware Always Wins)

Imagine you're a department head for a control unit. A classic hardware manager. Your performance agreement has one clear goal: reduce costs.

You find a way to make the memory chip slightly smaller. You save 50 cents per car. With 2 million cars, that's $1 million in savings. You get your bonus. Get promoted. You did your job.

Fast forward three years.

The car is on the market. The software team wants to roll out a new feature via over-the-air update. A feature customers would pay for.

But it doesn't work. Why?

The memory chip is full. Because the company chose the cheaper chip.

That $1 million saving today blocks $100 million in revenue tomorrow.

Here's the trap: Legacy automakers optimize everything for SOP (Start of Production). Hardware gets sized for today's functions. Not tomorrow's.

The core issue? Traditional automakers still work toward SOP. The moment when the product is finished.

But software is never finished.

When building a car today, automakers must think years ahead. They need extra computing power for tomorrow's features. Features that haven't been invented yet.

2) Middle Management Blocks Transformation

The second problem Michael calls the "Frozen Middle."

Here's how it works.

You're a middle manager at a legacy automaker. If the next big software project fails, you'll probably keep your job. If it succeeds, you won't become a millionaire.

So why take the risk?

There's no real incentive to push for radical change. Michael describes the pattern: Someone sits in meetings and keeps saying "That won't work because..." They block every proposal for three years straight. By then, they've transferred to another department or retired.

The transformation never happens. But they kept their job.

This isn't malicious. It's rational behavior in a broken system.

Taking risks? You might lose. Making no mistakes? You get rewarded.

The system was built for a different era. When "being ready for SOP" was the goal. Not continuous improvement.

3) "You Don't Even Know What Good Looks Like"

Here's what Michael means.

Decision-makers often don't understand what modern software development looks like. They come from the hardware world.

The problem: These managers are supposed to build great software products. But they don't know what great looks like. Because they've never seen it.

This has serious consequences for the teams.

Many developers at legacy automakers still work with outdated tools. Proprietary systems from another era.

Let me show you:

At software companies, code changes get tested automatically. The system gives instant feedback. Your code works or it breaks something. You know within seconds.

A person (or committee) must read and approve 500 lines of code. Manually. Feedback takes days or weeks.

This isn't just inefficient. Nobody can focus after reading 50 lines. It's also a massive recruiting problem.

Great software developers won't work with outdated methods. They want modern environments and tools.

Michael puts it this way: "Automotive could be the most attractive employer in the world. Writing code that makes a car drive autonomously? Much more exciting than e-commerce."

But top talent won't come without the right environment. And only leaders who understand software can create that environment.

We Still Have Time (But Not Much Longer)

Here's the good news.

EV adoption is slow in Europe and the US. That buys legacy automakers time.

Why? Because customers expect software in EVs. Like Tesla showed them. With combustion engines, they're still more forgiving.

But the buffer is shrinking fast.

China tells a different story. Customers there buy cars like consumer electronics. Like smartphones.

They expect digital features. They expect software. Legacy brands are increasingly seen as yesterday's technology.

What I learned from Michael

The biggest problem isn't missing software developers. It's not missing technology. The problem is 100% organizational and cultural.

Legacy automakers have organizations built for a hardware world. For predictable cycles. For "being ready for SOP." But software needs iterative development. Software is never finished.

This is a fundamental shift. Legacy automakers still operate like hardware companies.

Chinese newcomers have an advantage here. They start fresh. They build their organizations for the SDV era from day one.

Legacy automakers face a tougher challenge. They must transform while the business keeps running. Like rebuilding a ship at sea.

Here's what needs to change:

1. Change incentives: Hardware managers shouldn't get bonuses for cost cuts that kill future software capability

2. Redefine SOP: In software, Start of Production isn't the end. It's the beginning

3. Open career paths: Let top developers advance without becoming managers

Software now defines how cars drive, brake, and feel. It's time it also defines how we work.

PS: Want to go deeper? Download the free SDV Pulse Report from Michael and Thoughtworks here.

That's all for today.

Feel free to reply to this email with your thoughts.

Until next week,
Philipp

PS: If you find value in this newsletter, please share it with someone who might benefit.

After 10 years at Mercedes-Benz, I quit in 2020. In 2024, I started writing "Der Autopreneur". It became Germany's largest newsletter on automotive transformation. Now it’s also available in English.

LinkedIn | About me | German Version