In The Loop - Chapter 22: General Motors

28 min read

When the Great Recession hit, General Motors was ill prepared. After years of declining market share and soaring pension liabilities, it had a vulnerable balance sheet. As the recession deepened, it soon became clear the company wouldn’t survive without an intervention. Car sales plunged, losses soared and the balance sheet withered. When the books closed on 2008, the company had lost a staggering $30.8B and burned through $19.2B of its cash reserves. With cash dwindling, CEO Rick Wagoner went to Washington DC to seek a government bailout.

It came in the nick of time. On June 1, 2009 General Motors filed for bankruptcy in New York. The company had $80B in assets and $173B in liabilities. In a unique maneuver, its distressed assets were placed inside one entity, and its healthier assets were placed in another — the new General Motors. When the two parts exited bankruptcy proceedings in just 40 days as two separate companies, the US government invested $50B in the healthy GM for a 61 percent stake, in return for a promise that CEO Wagoner would step down. The assets of the distressed company were liquidated, including fourteen shuttered plants and Hummer, Saturn, Saab and Pontiac.

In the new General Motors, Chevy, Cadillac, GMC and Buick lived on. The company retained 3600 of the 6000 dealerships originally in its network, and 60,000 of the 80,000 employees it had prior to bankruptcy¹.

The emergence from bankruptcy wasn’t smooth. Consumer confidence had been shaken by the bankruptcy. General Motors stock had been delisted from the NYSE. The company was now owned by the US government. Decision making was top-heavy; employees up and down the line were afraid to make any mistakes. In the period from January 2009 to December 2010, GM employees had to adapt to four CEOs (Richard Wagoner, Fritz Henderson, Ed Whitacre and Dan Akerson).

Ed Whitacre, CEO for a year, described a cover-your-butt mentality. Decision making ground to a halt; everything had to be approved by everyone. Said Whitacre: “This contributed to a stall-and-delay mentality. Engineering projects had to be vetted by the board — why, I have no idea².”

In January 2014, Mary Barra took the CEO baton from Dan Akerson. 52 years old, she had worked her entire career at General Motors. She had started as an 18 year old co-op student, checking fenders and hoods to play for college tuition. After that she graduated from General Motors University with a degree in electrical engineering, and hired on at General Motors full time. After taking time to complete her Stanford MBA, she returned to GM and rose steadily in the engineering ranks, becoming VP Global Manufacturing Engineering in early 2008.

From that point forward, her tour of duty diversified. She became plant manager at the Detroit/Hamtramck Assembly Plant. Then, with the company in crisis, she was encouraged to take on the position of VP Global Human Resources, which she held from 2009 to 2011 — a tumultuous time for post-bankruptcy GM. After that she became Executive VP Global Product Development, where she oversaw design and worked to reduce the number of automobile platforms. In 2013 she added responsibility for the global supply chain. When she was announced as CEO, Barra made history. She had become the first ever female CEO of an automotive manufacturer.

And so began a transformation journey — one still very much underway. It all started with a bang. Barra had barely settled into her seat as CEO when GM was hit with a scandal. It was discovered that the ignition switch on many of its smaller cars was failing due to a design flaw. The flaw had been implicated in multiple deaths and injuries. Soon, Barra was testifying in front of Senate committees and mobilizing the company to deal with the crisis. In her first year as CEO, GM issued 84 safety recalls impacting over 30 million vehicles.

Ultimately, it would be determined that the faulty ignition switch had caused accidents resulting in 124 deaths and 275 injuries. In the end, GM paid out almost $7B in compensation and fines.

The recall was a defining moment for Barra. After an initial investigation she dismissed fifteen employees. Early on, she told analysts on a conference call, “We’re taking full responsibility for what happened³.” She then hired former attorney general Anton Valukas to investigate more deeply. He found that the design flaw was rooted in incompetence in the engineering function, made worse by a culture of bureaucracy and poor communication⁴.

Barra took dead aim at the bureaucracy that had stifled GM for years. She used the recall as a rallying cry for change. Employees were encouraged to voice concerns. Committees that had bottled up issues were challenged to deal with them head on. She launched a “Speak Up for Safety” hotline, placing safety at the core of the company culture. She challenged employees and managers to embrace transparency, to engage in open and honest dialogue and to be accountable for results. In a company as large as GM, she recognized that messages from the top needed to be simple so that everyone could understand what mattered most. All of these lessons were hard fruits of the ignition switch crisis. In a 2014 company-wide meeting, she said, “I never want to put this behind us. I want to put this painful experience permanently in our collective memories⁵.”

As the ignition switch scandal slowly shifted from crisis management mode to change catalyst, Barra turned her attention to GM’s future. GM had lost its edge. It needed to recapture competitiveness and inspire its workers. It wasn’t enough to win on design values. She decided that GM would need to become a technology company, embracing electric vehicles and autonomous driving. This would be the new path to market leadership.

She also knew she could never get there solely by working inside the legacy business. So to create a platform for digital transformation she embarked on a series of investments (including $500M into Lyft), product bets (Chevy Bolt) and acquisitions (including Cruise, an autonomous vehicle company; and Strobe, a leader in Lidar sensor technology). As the platform building blocks emerged, she took steps to protect them from the mothership. To pay for these new investments, she restructured the legacy business and invited external investment in Cruise.

Then in October 2017, she laid out her vision for the company: zero crashes, zero emissions, zero congestion. It has guided the company ever since.

As of this writing, General Motors is still a work in progress. The ultimate outcome of its generative and digital transformation will not be clear for a few more years. But it has committed itself a bold vision rooted in technology leadership. From all indications, it is well on its way. Perhaps its greatest asset is that its CEO is an In The Loop leader: ethically grounded, digitally literate, and a systems thinker.

Let’s consider the GM story in light of the eight core ideas in this book.

  1. In the face of continuous change, the fit systems enterprise relentlessly pursues a customer centric, world-improving Virtuous Vision

The roots of the new vision took time to emerge. Barra spent her first year caught up in the ignition switch scandal. But as she headed into her second year, she began to test the edges of a digital future. She invested in preparing the Chevy Bolt for market launch. Through its venture arm, GM Ventures, the company made a June 2015 investment in Proterra — a company developing zero emission battery-powered electric buses. It was the first glimmer of bigger things to come.

2016 was a big year in GM’s transformation. In January, she made her first big bet — a $500M investment in Lyft. GM president Dan Amman took a seat on its board. Two months later, she acquired Cruise Automation for $581M plus incentive based compensation. Cruise would later become the centerpiece of GM’s digital transformation. The Chevy Bolt was released in the Fall, the first all-electric vehicle priced under $40,000 to exceed a range of 200 miles per charge. The Bolt had been Barra’s from the beginning. A team in South Korea began development of the vehicle in 2012, when Barra was EVP of Global Product Development. Other investments in 2016 included SolidEnergy, a company working on next generation batteries, and Alphabet Energy, a company with technology to turn heat (such as from engines) into electric power.

In June of 2016, after an impactful trip to Silicon Valley, Barra posted a LinkedIn article in which she shared her reflections on leadership. Reading this post, one senses a new company vision undergoing active formation.

“Where do we start? For me, at GM, it starts with customers…. And no matter what business you go into, you only win when your customer says you win….

As a visionary leader, you should be thinking about more than just the next quarter. You should also be thinking about the next decade… and what your company’s reputation and place in the world will be after 40 quarterly results.

More and more, today’s employees want to be connected to a broader purpose, a higher calling. They want their companies and institutions to make the world a better place. I believe we can do both. I believe we are required to do both. I believe it’s up to leaders to set the tone, create the vision, and inspire the behaviors that allow our organizations to best serve society⁶.”

Sixteen months later, on October 3, 2017 she announced a new vision for General Motors. Her post on LinkedIn that day reads in part:

We are in the midst of a transportation revolution….The good news is that our generation has the ambition, the talent and the technology to realize the safer, better and more sustainable world we want. General Motors has committed itself to leading the way toward this future, guided by our vision of zero crashes, zero emissions and zero congestion.

We see a future of zero crashes where, instead of trying to protect passengers during and after a crash, we harness new technologies to prevent crashes from happening in the first place. With safety as General Motors’ number one priority, we’ll keep innovating…Even as we continue to deliver the best fuel economy in the vehicles our customers love to drive today, we’ll move relentlessly and irreversibly to a zero emissions future. No more gas. No more diesel. No more carbon emissions…

And we see a world of zero congestion where new technologies allow us to travel smoothly, safely and more quickly to our destinations. No more sitting in traffic. No more wasted time, fuel and money. And just maybe, no more road rage…

Zero crashes — so we save lives.

Zero emissions — so our children can inherit a healthier planet.

Zero congestion — so our customers get back a precious commodity: time.

This is GM’s vision, and this is the future we’re going to deliver to move humanity forward⁷.”

That’s a customer-centric, virtuous vision. A post on makes clear that it’s taken seriously: “…our ambitious plan of zero crashes, zero emissions and zero congestion drives every decision we make⁸.”

2. To do this, the enterprise must advance two imperatives: to be Generative, and to be Adaptive. Being generative (continuously creating new value) is always the first priority; being adaptive (increasing resilience, scalability and efficiency) is second

For many years, GM didn’t focus much on its generative imperative. As the company continued its long market share slide, company leaders placed primary attention on the least important adaptive priority: efficiency. The 2009 bankruptcy made it abundantly clear that the company would never be able to save its way to prosperity. It need to innovate. When Barra took over, she put in place the mechanisms to do just that.

A company pursues its generative imperative by actions inside three systems: product discovery, product management and corporate development. For Barra’s GM, much of its product discovery and corporate development work has been accomplished through investments by GM Ventures — in autonomous vehicle technology, batteries, sensors and other enabling technologies. The Chevy Bolt is an exception, initially developed by a GM team in South Korea.

GM’s most important investment has been in Cruise Automation. This San Francisco based company, founded in 2013 by Kyle Vogt and Dan Kan, had about 40 employees when GM wrote a $581M check (plus incentive based payments) to buy the company in March of 2016. By Fall of that year, the company was already up to 100 employees. By June of 2017, with 200 employees, GM had announced that it would invest more into the company and grow to 1,163 full-time employees by 2023. Just under a year later, in May 2018, it was announced that SoftBank would invest $2.25B in the company, along with another $1.1B from GM. Then just five months later, Honda announced it would invest $750M into the company, followed by another $2B over the next twelve years. The investment valued the company at $19B⁹.

In late 2017, GM announced it had acquired Strobe, an eleven-person company working on next generation Lidar technology — a key component in the sensor system for autonomous vehicles. The company is also based in San Francisco, smoothing its integration into Cruise.

Today, 180 Chevy Bolts autonomously roam the streets of San Francisco; Scottsdale, Arizona and Detroit — stacked to the gills with autonomous vehicle technology. Reams of data from these test vehicles stream into Cruise to power its optimization efforts. In San Francisco, Cruise employees use its test vehicles for ridesharing around the city. GM’s investment in Lyft provided insights into the ridesharing market, and Cruise’s nascent ridesharing program is building on those early lessons. GM has targeted late 2019 to launch its first fully autonomous vehicle into the market.

Cruise has posted steady reliability gains since the GM investment. The State of California requires reporting from any company that deploys autonomous vehicles on California roads. A key measure is the disengagement rate. This is the frequency with which the safety driver needs to take over the car (disengaging the autonomous system) in order to respond to a traffic event. The performance of GM Cruise in 2018 was second only to Google’s Waymo, with a disengagement once every 5,205 miles — 321% better than the year prior. Much of the company’s testing was in San Francisco, one of the most complex traffic environments in the country. As Cruise co-founder Kyle Vogt said in a 2017 blog, “Every minute of testing in San Francisco is about as valuable as an hour of testing in the suburbs¹⁰.”

GM’s investments in the Chevy Bolt and Cruise Automation are multi-billion dollar bets that bring the company’s stated vision to life. The Bolt presages a day when all GM vehicles are electric — Zero Emissions. And Cruise presages a day when autonomous driving virtually eliminates the 1.5 million deaths per year worldwide from automobile crashes — Zero Crashes. It also points to a future world of intelligent traffic coordination — Zero Congestion. Zero, Zero, Zero.

It is not yet entirely clear how the autonomous vehicle market will evolve, nor how quickly the shift to electric vehicles will happen — but GM is now well positioned to be a leader in both of these emerging markets.

As important as it is for a company to be generative, that’s not sufficient. It must also be adaptive. Barra has worked hard to break down the bureaucracy of the company, so as to increase feedback loops and ensure the right information gets to the right people every time. She has worked to build a more nimble culture, characterized by transparent communications and efficient decision making. These steps have increased company resilience.

But she has also made difficult decisions to restructure the company, creating efficiencies so that money can be invested into innovation. In 2017, GM exited Europe after twenty years of losses, selling its Opel and Vauxhall divisions. That same year, it also phased Chevrolet out of India and South Africa. Then in November 2018, the company announced a major restructuring, reducing expenses by $4.5B and capital expenditures by another $1.5B — $6B total. It closed five plants, consolidated propulsion and engineering teams, took steps towards consolidating vehicle architectures, changed design practices to increase the number of components that can be shared across multiple vehicle product lines, and reduced the global workforce — including a twenty-five percent reduction in management positions.

Barra said at the time, “The actions we are taking today continue our transformation to be highly agile, resilient and profitable, while giving us the flexibility to invest in the future… We recognize the need to stay in front of changing market conditions and customer preferences to position our company for long-term success… These actions will increase the long-term profit and cash generation potential of the company and improve resilience through the cycle¹¹.”

That’s the adaptive imperative in action.

3. When conceiving of the enterprise, the Systems View is primary; the functional view must be secondary

Barra’s path to the top began on the shop floor. Along the way she spent time in most of the critical functions of the business. Her role as head of global human resources during the years right after bankruptcy provided her with a systemic view of the company.

Since becoming CEO, Barra’s systems view is reflected in her decisions. One example is the vertical integration path she has taken with the Chevy Bolt and the Cruise acquisition. In the same way that Apple vertically integrated its iPhone, delivering users a more complete phone experience than is possible with Android phones, GM is gaining control of the building blocks of electrical and autonomous driving. This began with the Cruise acquisition, but was advanced by the acquisition of Strobe — the Lidar technology company. Now it has control over the building of the car itself, the autonomous driving brain (Cruise) and its eyes (Strobe). It has even begun experimenting with the service layer, both in its Lyft investment and in its own rideshare service for Cruise employees, called Cruise Anywhere. By owning all the building blocks, it is able to build comprehensive system feedback loops into the technology. It also gains more control over the production cycle for components. Now components can be built to match GM’s specifications, not the other way around. And GM now benefits from all the economic value of each component, versus having to share that value with a supplier¹².

4. The enterprise is comprised of Operating Systems (such as the revenue engine system and the product management system) that create results; and Meta Systems (such as the governance system and culture system) that exist to elevate the performance of operating systems

The operating systems of a company as large and diverse as GM don’t lend themselves to adequate review in a single section of one chapter of a book. Beyond the product discovery and product management systems mentioned earlier, GM’s revenue engine system includes four major subsystems: its global marketing engine (brand marketing), its national marketing engine (focused on promotions), its regional marketing engine (for regional promotions) and its dealership network. These subsystems exhibit significant variations in different parts of the globe — such as in China, where GM has ten joint ventures. Its production system keeps vehicles rolling off the production line. And of course GM features a comprehensive global accounting system and HR system.

I identified earlier in this chapter the role GM’s corporate development system has played in rebooting its product discovery system and advancing its product management system. A continuing series of investments by GM Ventures, along with the acquisitions of Cruise and Strobe, have combined help GM advance its generative imperative.

But perhaps less obvious is the financial engineering required to pull all this off. GM has managed its cash management system with a high degree of fidelity, enabling it to invest billions of dollars in its digital transformation. Take, for instance, the decision to accept outside investment in Cruise. This has enabled the company to raise well over $5B in committed funding it would otherwise not have been able to raise. It has done so while retaining more than 50% ownership in the company. Similarly, it has made other smart bets with its cash. At the share price as of the day of this writing, GM’s $500M investment in Lyft is now worth well over $1B. And its restructuring efforts have saved over $6B, freeing up cash flow for transformation investments.

But these are operating systems. Perhaps the biggest systemic impact Barra has made has been on the company’s meta systems.

The Governance System

A company’s governance system is made up of its executive team, CEO and board. Given that Cruise Automation is a majority-owned subsidiary, its leadership is also part of GM’s governance system.

Barra has made a series of moves to build the team of her choosing. GM’s president, Mark Reuss, was promoted to the role in January 2019. His background in global operations and product development equips him for the traditional parts of the job, but it was the role he has played in the transition to autonomous and electric vehicles that Barra cited in the announcement. That same month, Dan Amman was moved out of the GM president role and became CEO of Cruise Automation. Amman had been overseeing the company since its acquisition, protecting it from mother ship incursions. With Amman CEO of the acquired company and its former CEO Kyle Vogt (now CTO) retained, the leadership is in place to ensure the right balance of isolation and integration at GM-owned Cruise.

Barra’s senior team appointments reflect the company zero, zero, zero vision. There is a healthy mix of auto manufacturing veterans and executives with experience in autonomous driving, electric vehicles and modern technology — and a mix of GM veterans and new blood. The company’s five executive vice presidents (Gerald Johnson, global manufacturing; Matt Tsien, GM China; Bary Engle, GM Americas; Dhivya Suryadevara, CFO; and Craig Glidden, General Counsel) all were hired concurrent with or since Barra’s appointment to the CEO role. Of its nineteen senior vice presidents, country presidents and various “chiefs” (diversity, compliance, accounting, marketing, data, technology and information officers), just five preceded Barra in their roles. Many of her key moves have been made in the past two years. The occupants of eleven out of the twenty-four positions that sit between Barra and the SVP level were appointed in 2018 or 2019.

Executive team composition and alignment is not the only priority. It’s equally important to build a strong and aligned board.

“It’s a very involved board,” said Barra in an interview in January 2019. “We’ve got to face the brutal facts and understand what’s happening and seize opportunities as opposed — and lead disruption instead of being disrupted. As we start to think of something new or we see an opportunity, we have a discussion. And then we’ll bring it up again at the next board meeting; or if things are moving so quickly, we all jump on a call.” She described board membership as diverse, but indicated that board composition is regularly assessed. “We have a very robust process where, every year, we look at what are the skill sets we need not for today but for the future. And that’s what drives us as we do board refresh¹³.”

Despite a stock price that has proven stubbornly immobile since Barra’s arrival, the public evidence indicates a supportive board that stands behind her key decisions — especially the company’s long-term investments into autonomous driving and electric vehicles.

The Strategy / Planning / Architecture System

It’s not easy to coordinate work across three global regions, four vehicle brands, a complex dealer network and the many functions required to manufacture cars and trucks. It’s even harder when you are seeking to transform the product itself — shifting from internal combustion to electric vehicles, and from human-driven to self-driving vehicles. You have to balance now, near and far, and coordinate across domains.

But Barra is building the planning disciplines and communications architecture to get there. To gain an idea of the level of coordination between strategy, planning and architectures (such as organization architecture, technical architecture and financial planning architecture), consider the following sequence of events:

  • Late 2014-early 2015: Rumors emerge that GM will launch Super Cruise, its new autonomous vehicle technology, in the Cadillac CT6, in 2016 or 2017
  • August 2015: Cem Saraydar, Director of GM’s Electrical and Control Systems Research Lab indicates Super Cruise will allow drivers to fully take their hands off the wheel and feet off the pedals — ahead of all other competitors
  • January 2016: GM invests $500M in Lyft; Dan Amman appointed to Lyft’s board
  • March 2016: GM buys Cruise Automation; CEO Kyle Vogt will report to Dan Amman
  • Super Cruise is launched on 2018 CT6 Sedan, allowing hands and feet to be free for driving on some highways
  • October 2017: GM acquires Strobe, a tech startup building next generation Lidar sensor technology
  • May 2018: Softbank invests $2.25B in Cruise
  • May 2018: Software update to Super Cruise to address sensor issues in which sunlight tricks the system (a problem Lidar technology can solve)
  • October 2018: Honda invests $750M in Cruise
  • January 2019: Dan Amman announced as CEO of Cruise Automation
  • GM announces that Super Cruise will be available on all Cadillacs by 2020, and will extend to other GM brands after that

You get the sense that Barra and the GM executive team are developing a sophisticated strategy / planning / architecture system.

Culture System

Barra leveraged the ignition switch scandal to attack GM’s bureaucratic, cover-your-butt culture. She clarified the company’s vision, and challenged leaders to break down barriers to communication — especially bottom-up and cross-functional communication. She has delegated responsibility and increased accountability for results. But beyond that, she has worked with executives, managers and employees to build a culture that is safe, engaged, diverse, enabled, aligned and driven.

Here is how she describes it:

“My ideal workplace is built on that “can do” attitude and is strengthened by the relationships we build inside and outside our organizations. In order to accomplish this, leadership and teams must be:

  • Safe — a great workplace makes safety a foundational commitment for employees and the customers they serve.
  • Engaged — people are excited about their work, proud of their organization and eager to make their work environment better. They feel valued for doing meaningful work. They trust their leaders, and their leaders trust and empower them.
  • Diverse — a true team that is diverse, inclusive and reflective of the customers it serves. Diversity, in all of its many dimensions, makes workplaces stronger, including diversity of thought and experience.
  • Enabled — every employee is empowered to contribute to his or her full potential. They are encouraged to contribute and speak up when something needs fixing or improving.
  • Aligned — employees are committed to the organization’s core values and behaviors. They work together to win.
  • Driven — people are committed to excellence. They lead by example and know that integrity, accountability and results are expected and rewarded.

A great workplace creates an environment — and gives us the freedom — to pursue the art of the possible¹⁴.”

DataOps System

Since his hiring in 2012, Randy Mott, Chief Information Officer, has led the consolidation formerly outsourced IT resources into the company. Data has been at the heart of the transformation. For example, Mott’s team built a comprehensive data system called Maxis, essentially a search engine for all its proprietary and third party data. Now the company can do such things as to identify the profitability of every vehicle sold — something it could not previously do. The 2017 hiring of Charles Thomas as Chief Data Officer marked another key step in creating a unified data infrastructure across GM. At a customer analytics conference held at Wharton, Thomas said the following:

“You have to have singular accountability for data and analytics because you’ve got privacy, you’ve got all these other issues around legislation, around hacking and concerns,” explained Thomas. “Of course, I work with the Chief Information & Security Officer and Chief Privacy Officer. But I’m the singular tip of the spear for data, regardless of what kind of data it is¹⁵.”

As the Maxis project indicates, GM clearly has a commitment to put their data to work in decision making. And it’s working hard to ensure the data is properly protected and managed with proper permissions. The company is still far from completing the work to push data driven decision making and democratized data access into all the nodes of its business. But it is heading in that direction.

Engineering System

The engineering system at Cruise Automation is finely tuned. It could not be otherwise, given what it does. As to GM itself, its engineering system is at an earlier stage of maturity. It was only in 2012 that GM began to move from fully outsourced IT to insourced development and operations. Since that time, it has built out a technical environment that serves all of its operating systems, including its products.

But it’s clear that GM’s digital transformation journey is still in its early stages. Mott has brought technology development in house and has built enterprise data centers. But there is some evidence that his transformation path has not taken full advantage of modern technical capabilities. For instance, Mott is on record as being skeptical of public cloud-based systems. He has built a private cloud in two enterprise data centers filled with GM hardware. This means GM has had to buy and maintain hardware sufficient to handle its computing demand peaks, versus in the public cloud where you can move up and down as required and only pay for the compute power you use.

In an interview in 2014 with Information Week, Mott stated that an embrace of cloud software and infrastructure would be “a strategy of how to I get even, not how do I get ahead¹⁶.” This is perplexing, given the proven benefits of cloud-based architectures for companies worldwide. His technology vendor choices (including Oracle and SAP) also hint at a less than fully modern vision for technology at GM. One is left to wonder whether GM corporate has built the technical capacity to take full advantage of new technologies such as machine learning and AI, especially those that depend on streaming data.

Contrast that with Cruise Automation, an enterprise-scale user of Google Cloud. In a post on, technology lead manager Karl Isenberg hints at the company’s technical sophistication. “Cruise PaaS spans multiple Google Kubernetes Engine (GKE) clusters in multiple Google Cloud Provider (GCP) environments and projects, with a bunch of addons to increase the functionality of GKE and make it work on our private hybrid-cloud network,” he said.

GM has certainly embraced many emerging technologies. Over 6 million vehicles are connected to OnStar via 4G LTE technology. Today, the average new GM car is powered by over 100 million lines of code — more than a Boeing 747 jetliner. As a result of the digitization of the vehicle, GM’s data centers now process over 88 million gigabytes of data a year.¹⁷

Still, we must assume that at GM corporate, the engineering system is a work in progress. The company is not fully utilizing the benefits of the public cloud. Unlike global enterprise counterparts such as Starbucks, Netflix and Amazon, online searches yield no evidence that GM has organized its technical teams into autonomous cross-functional domain teams, nor that they are pursuing disciplined agile delivery methods. Perhaps in time the lessons from Cruise will begin to infiltrate the mothership.

5. The fit systems enterprise features Domain Driven Design — both in its technology design and in its organizational design. All systems in the enterprise are made up of domains; these domains integrate people, workflows, technology and money flows to achieve business outcome objectives.

If you lead a large established enterprise with a formidable but threatened legacy business, it is on you to carefully assess how much change to impose. Certainly there is a need to embrace the new, but you also must tend to today. The former requires agility. The latter requires stability.

General Motors has been confronting this challenge since Barra became CEO. Within the legacy business, change has been introduced incrementally — starting at the top and working down. Through restructuring, reorganization and senior team changes, Barra has put in place agents of change who can push the edges of agility while retaining sufficient stability.

But we can see in GM’s acquisition and integration of Cruise the unfolding of a much more bold plan of transformation. Barra understood from the outset that the most important first priority with the Cruise acquisition was to protect it. To do that, she needed to isolate Cruise from the rest of GM. That was the mission she gave to Dan Amman when he took on responsibility for Cruise after the acquisition. Michael Arena, Chief Talent Officer at GM, described it this way in an interview with Deloitte Insights:

“You know Cruise Automation is our self-driving technology bet. But it’s completely partitioned off from the organization. You literally cannot get into Cruise. I cannot get into Cruise without actually going through a gatekeeper. That’s by design. That’s 100 percent by design.

The gatekeepers are the bridge people that keep that growth path connected to the organization but disconnected enough so that it can move fast. What we know is it can move 20 times faster in the design that it’s in right now. But at some point that will be a problem, and this is why incubators and accelerators fail — because at some point that needs to become the new core. So, the bridge people are the most important people inside of the organization now because they keep this growth entity that we treat like a startup, we incent like a startup, we hire there separately. Everything is separate. It’s a whole distinct set of HR practices. We evaluate differently. We have different talent reviews with those folks from the core¹⁸.”

Dan Amman was the key bridge player in the case of Cruise. One of the critical early decisions Amman made was to create an equity-like incentive structure for top management at Cruise. It was unlike anything offered to executives or managers at GM. He knew that such an incentive structure would be required if he wanted to avoid losing the Cruise top team to a new hot startup.

Arena went on:

“The core is handled just like many large organizations, and the bridge people are in the middle. And those bridge people came from the core on purpose because they have credibility with the core. We’re thinking about bringing their social capital over with them as we put them in the bridge. The reason for that is the core trusts them, and they’ve been positioned with the growth path in this particular case. We now are replicating this many, many times. With that growth path, those people are trusted because they become almost like the librarian to the 105 years’ worth of legacy knowledge to do things like durability testing. To do things like how do you work with public policy to get this thing adopted? How do you deal with the insurance companies? All of those other things that are critical to a skeletal operation. In the end, you’re bimodal, if not trimodal, from an HR practice standpoint¹⁹.”

Today, with over 1200 employees, Cruise is comprised of many systems and domains. Its infrastructure team alone has 330 employees and its autonomy software team has 150; by all indications these teams are cross-functionally organized into domain teams²⁰.

Mo ElShenawy is VP engineering at Cruise. In a blog post, he describes the organization structure:

“For instance, our Senior Software Engineers (more than 50% of our engineering group) understand their team’s software very well. They are great at designing solutions, writing model code, and deploying high quality work consistently and autonomously. They work directly with their stakeholders to deliver the right solutions….

On top of the above expectations, our Staff Engineers (around 5% of our engineering group) are considered the technical leaders across their larger group (Sr. Mgr level or Directors). They initiate large projects with more sophisticated architecture and break it down to the right components so other engineers can utilize it effectively. Staff Engineers also have a wide system view across the org, and they work across teams to coordinate cross-system architectures….

Our Principal Engineers are the few technical leaders responsible for the coherence of the engineering community at Cruise. They are often domain leaders helping advance the technology for the greater good of the industry. They guide the overall technical strategy based on a deep understanding of the domain and industry context…²¹”

Kyle Vogt remained CEO of Cruise from its acquisition in March 2016 until Amman took over in 2019 — a period of over two and a half years. This enabled Amman to build trust with the Cruise team, so that by the time Amman became CEO, Vogt felt comfortable staying in place as CTO. Now the foundation for transformational change is in place. Cruise has retained its independence in culture and practices, but a bridge builder is at the top — one that is trusted by the legacy business. And on the GM side, new bridge builders are being put in place to receive and evangelize the ideas emerging out of Cruise — such as Patricia Fletcher, VP innovation, and Doug Parks VP autonomous and electric vehicle execution. As technology and best practices emerge from Cruise over time, these bridge players are positioned to evangelize change within legacy GM.

Later in his interview with Deloitte, Arena addresses how change happens inside large organizations. He notes that pockets of employees (domain teams?) are needed to discover new ideas, develop them, diffuse them and then disrupt existing structures and systems. Each of these teams involves different types of people.

“The more sophisticated organizations are thinking about different social arrangements, and those are the four D’s that I talked about in the book. There are times where we need to be discovering ideas, and that requires one set of design and/or arrangements (brokers). There are times where we need to be developing those ideas, and that requires a different set of arrangements for a defined period of time (connectors). There are times where we need to take those ideas out of those entrepreneurial pockets and diffuse or scale them across the organization. That requires a different set of connections (energizers). And then ultimately those three will add up to disrupting the current structures and operational systems (challengers). These network roles are the way to enable that²².”

6. In the fit systems enterprise, data access is democratized; there are Feedback Loops Everywhere

To get a sense for the sophistication of Cruise’s technical feedback loops, it is helpful to peek into their simulation system — the system that conducts simulations of diverse driving scenarios so as to refine vehicle response. In a blog on, Rick Fulton, senior engineering manager at Cruise, said the following:

“We face some unusual challenges in storing simulation data. We produce gigabytes of simulation data per second that need to be available within minutes to allow engineers to perform quick analyses after running tests. We’ve also been dramatically increasing the number of simulations we run, so we need an analytics database that can effortlessly scale.

To understand why there is so much data and why it’s so complex, we should look at what makes simulation testing harder than traditional software regression testing.

First, simulations do not necessarily output binary results. For example, in a given version of the code an x% decrease in one metric may be acceptable if accompanied by a y% increase in another metric. Because of the number of metrics required per sim, each simulation generates upwards of a gigabyte of result data²³.”

As to its organizational feedback loops, Cruise is transitioning from a small “move fast and break things” startup to a company building the rigor necessary to create safe autonomous vehicles at scale. Said Fulton: “This analysis platform and underlying data infrastructure have fundamentally reshaped AV development workflow at Cruise. Now, AV engineering projects always include metric targets and visualizations built with these tools²⁴.”

At GM, the shift from a siloed, “cover-your-butt”, under-communicating company is still underway. Barra has built a strong safety culture, which has required increased collaboration. And the Maxis system has enriched the data and analysis that can be brought into decision making.

7. The fit systems enterprise exhibits rising Digital Leverage: its systems are built via reactive microservices architecture; its data infrastructure is modern and cloud-based; and its technical domain teams pursue disciplined agile delivery methods to build powerful solutions

Cruise has built its systems via reactive microservices architecture, on the cloud, leveraging modern data infrastructure. Fulton shows how the autonomous system works at a high level:

He then shows and describes the Cruise simulation pipeline as follows:

“The pipeline for Simulation results is the following:

  1. In a typical run, our simulations get scheduled and run through our simulation service.
  2. A graph compute engine transforms the raw simulation output into Avro tables (our chosen data serialization protocol).
  3. These tables get stored in Google Cloud Storage (GCS)
  4. After this upload to GCS is complete, our simulation service notifies a simple ingestion service via pubsub that there are new simulation results to pull.
  5. The ingestion service retrieves the latest data, examines the tables for any new added metrics (new columns), makes the necessary changes to the BigQuery schema.
  6. The ingestion service then uses the streaming insert API to add the simulation results to BigQuery.
  7. At this point, the data is available to various consumers, such as Jupyter notebooks and Business Intelligence (BI) tools like Looker and Tableau²⁵.”

At GM, its level of digital leverage is much lower — but it is rising. The pockets of digital transformation are expanding. Ahmed Mahmoud, global chief information officer for manufacturing and supply chain at GM, was interviewed in an article in published in September 2018. In this article he described the growing use of microservices, rich APIs and service virtualization — indicators of modern technology adoption²⁶. Similarly, GM CTO Jon Lauckner envisions the continuing adoption of cutting edge technologies, especially in the realm of connectivity. He notes that OnStar was an early adopter of 4G LTE, and that the company has begun to assess the impact 5G will have on future development²⁷.

Nonetheless, one gets the sense that the digital transformation journey is still early at GM.

8. The fit systems enterprise exhibits a key foundational asset: “In The Loop Leaders.” They are ethically grounded, digitally literate systems thinkers. These leaders focus most of their attention on advancing a virtuous vision, improving the meta systems, ensuring domain team autonomy and increasing the density of 10Xers in high variation domains

Mary Barra is an In The Loop leader. Her vision is ethically grounded. A series of decisions, such as the investment in Lyft and acquisitions of Cruise and Strobe, and support for the digital transformation of GM reveal that she is digitally literate. And she is a systems thinker. Her 33 year career at GM gave her exposure to a diverse array of roles within GM prior to becoming CEO, so she knows the company through and through. Her placement of bridge players like Dan Amman, Patricia Fletcher and Doug Parks in key bridging roles underscores her understanding of the system dynamics and archetype risks that could exist between old and new. And her capacity to balance now, near and far, as seen in her restructuring actions to address current performance while simultaneously investing aggressively in an uncertain future are further indications of a sophisticated systems view.

Barra has built a healthy governance system, strategy / planning / architecture system and culture system. She is working to elevate the DataOps and engineering systems, though the progress there is more uneven. And she has certainly worked to elevate the quality of her leadership team, making a series of promotions, hires and replacements — especially since 2017.

As shown in a speech she gave at Wharton in 2018, her guidance to leaders emphasizes eight key points (my commentary in parentheses):

  1. Ask for feedback from your staff (increase feedback loops)
  2. Meet to discuss, not disseminate information (efficient use of leadership time)
  3. Simplify your message (create alignment)
  4. Effect change (business outcomes, not process outputs)
  5. Know the business (systems view)
  6. Win both hearts and minds (a virtuous vision)
  7. Align on values (build the culture)
  8. Lead culture with behaviors (in the culture system, actions are more important than words)²⁸

She has modeled and messaged a new approach to leadership at GM, and other leaders — like Dan Amman, Mark Reuss, Michael Arena and others — have bought in. Now they are sharing the transformation load.


General Motors has many miles to go on its transformation journey. The stock market has yet to reward GM’s progress. Investors know that economic downturns are particularly painful for auto manufacturers, and GM is sure to face one in the not too distant future. But under Barra’s leadership, the work of transformation is steadily advancing. As long as the company’s board keeps faith in the journey, GM will continue its march towards a future of electric vehicles and autonomous driving.

If someday we find ourselves in a world where streets and highways emit zero emissions from vehicles, zero crashes and zero congestion, then we will be able to thank Mary Barra and General Motors for helping make it happen.

To view all chapters go here.

If you would like more CEO insights into scaling your revenue engine and building a high-growth tech company, please visit us at, and follow us on LinkedIn, Twitter, and YouTube.


  1. Bigman, Dan. “How General Motors Was Really Saved: The Untold True Story Of The Most Important Bankruptcy In U.S. History.” Forbes. Forbes Magazine, November 6, 2013.
  2. Taylor, Alex. “10 Big Problems for General Motors.” Fortune. Fortune, February 14, 2013.
  3. Vlasic, Bill, and Hilary Stout. “G.M. Results Show Financial Hit of Recalls.” The New York Times. The New York Times, July 24, 2014.
  4. Feloni, Richard. “GM CEO Mary Barra Said the Recall Crisis of 2014 Forever Changed Her Leadership Style.” Business Insider. Business Insider, November 14, 2018.
  5. Ibid.
  6. Barra, Mary. “Leadership Lessons from The Valley: I Didn’t Know What I Didn’t Know.” LinkedIn, June 13, 2016.
  7. Barra, Mary. “Zero Crashes, Zero Emissions, Zero Congestion.” LinkedIn, October 3, 2017.
  8. “What We Really Build Is Relationships, Opportunities, Hope and Memories.” General Motors, n.d.
  9. “Cruise Automation.” Wikipedia. Wikimedia Foundation, July 30, 2019.
  10. Vogt, Kyle. “Why Testing Self-Driving Cars in SF Is Challenging but Necessary.” Medium. Cruise, October 10, 2017.
  11. “General Motors Accelerates Transformation.”, November 26, 2018.
  12. Bhuiyan, Johana. “Why GM Is Vertically Integrating as It Moves Deeper into Making Self-Driving Cars.” Vox. Vox, October 9, 2017.
  13. LaReau, Jamie L. “Mary Barra Says GM Board Is Closely Involved in Day-to-Day Decisions.” Detroit Free Press. Detroit Free Press, January 17, 2019.
  14. Barra, Mary. “Six Things That Make a Good Workplace Great.” LinkedIn, July 5, 2016.
  15. “Wharton Customer Analytics Conference.” Wharton Events, n.d.
  16. Preston, Rob. “GM’s Randy Mott: What I Believe…” InformationWeek, March 27, 2014.
  17. Wayland, Michael. “GM Plots next Phase of IT Overhaul.” Automotive News, September 18, 2017.
  18. McDowell, Tiffany, and Anh Nguyen Phillips. “The Power of Networks.” Deloitte Insights, January 16, 2019.
  19. Ibid.
  20. Efrati, Amir, and Mike Sullivan. “Org Chart: Cruise.” The Information, February 19, 2019.
  21. ElShenawy, Mo. “An Engineering Philosophy for Individual and Company Growth.” Medium. Cruise, July 9, 2019.
  22. McDowell, Tiffany, and Anh Nguyen Phillips. “The Power of Networks.” Deloitte Insights, January 16, 2019.
  23. Fulton, Rick. “Data Warehousing for AV Simulation Analysis.” Medium. Cruise, May 11, 2019.
  24. Ibid.
  25. Ibid.
  26. Baptista, Alan. “GM: Innovating Through APIs, Insourcing, and Service Virtualization — DZone Integration.”, September 23, 2018.
  27. DeBord, Matthew. “GM VC Boss: ‘The Automotive Ecosystem Is Changing’.” Business Insider. Business Insider, February 5, 2016.
  28. Barra, Mary. The Wharton School of the University of Pennsylvania. Philadelphia, May 9, 2018.
Tom Mohr

Leave a Reply

Your email address will not be published. Required fields are marked *