The auto industry's struggle to recruit and retain software engineering talent has been a recurring theme in the industry trade press for at least a decade, with the urgency of the commentary ebbing and flowing as the macro tech-employment cycle has gone through its various phases. The problem has been real throughout that period, but it has been manageable — the auto industry has built up internal software capability through a combination of in-house hiring, partnerships with external software companies, and strategic acquisitions of smaller software-focused organizations.
The conditions that have made the talent problem manageable are changing in ways that are going to make the next several years substantially harder for the industry. The software-intensive product roadmaps the major automakers have committed to require talent at scales and skill levels that exceed what the industry has historically recruited successfully. The competitive environment for software talent has reset in ways that put the industry at a structural disadvantage. And the internal organizational dynamics that have shaped the industry's software-engineering function have not yet adjusted to match the strategic priorities the industry has placed on software capability.
What the industry has been doing for software talent
The auto industry's approach to building software capability over the past decade has involved several parallel strategies. The major manufacturers have built up in-house software engineering organizations, with the engineering functions at the major automakers now collectively employing tens of thousands of software engineers across vehicle-software, connected-services, and corporate-software functions.
The geographical distribution of the in-house software engineering has expanded substantially. The traditional engineering centers in Detroit, Stuttgart, Wolfsburg, and other historical OEM headquarters cities have been supplemented with software-focused locations in major tech hubs — Silicon Valley offices for most of the major automakers, Seattle and Austin software centers, European software centers in cities other than the traditional engineering centers, and Asian software centers in cities like Singapore and Bangalore.
The acquisition strategy has been an important supplement to organic hiring. The major automakers have acquired companies across the software stack — autonomous-vehicle software developers, connected-services platforms, AI and machine-learning organizations, and various other specialized software capabilities. The acquired companies have generally been integrated into the broader OEM software organizations, with the talent retained where possible and the technology incorporated into the OEM platforms.
The partnership strategy has been the third leg of the talent approach. The collaborations between OEMs and external software companies — Google, Apple, Amazon, Microsoft, NVIDIA, Qualcomm, and various others — have provided OEM access to software capabilities that the OEMs could not have built independently at the speed the product roadmaps require.
What the strategic priorities now require
The software priorities that the major automakers have committed to for the latter half of this decade and beyond involve software development at scales and complexity that go beyond what the industry has historically delivered. The transition to software-defined vehicles — vehicles whose functionality is principally determined by software rather than by mechanical and electrical hardware — involves software development across vehicle systems, cloud infrastructure, mobile applications, and back-end services that collectively exceed the software content of any vehicles the industry has previously shipped.
The advanced driver assistance and eventual autonomous-driving systems require software development at safety-critical levels with verification and validation requirements that are substantially more demanding than typical commercial software. The connected-services platforms require ongoing development and operation at scales comparable to mainstream consumer technology platforms. The over-the-air update capabilities require the kind of release management, security, and operational discipline that the consumer-tech industry has spent decades developing and that the auto industry is still working through.
The cumulative software-talent requirement implied by these priorities is substantial. The internal estimates that have leaked into industry commentary suggest that the major automakers collectively will need to roughly double their current software-engineering headcount over the next several years to deliver on the announced product roadmaps. The doubling is from a starting point that has already been the result of substantial expansion over the past decade.
What the talent market is actually doing
The talent market for the categories of software engineers the auto industry needs has been shaped by the broader tech-industry employment cycle. The post-2022 adjustment in the consumer-tech sector, with the layoffs and reorganizations at the major tech companies, briefly produced a more accessible recruiting environment for the auto industry. Several major OEMs took advantage of the window to expand their software hiring at relatively favorable terms.
The window has been closing. The tech industry has stabilized and is again competing aggressively for the specific software-engineering skill sets the auto industry needs. The compensation levels at the major tech companies have continued to escalate, with total-compensation packages for senior software engineers at the leading tech companies substantially exceeding what OEMs typically offer. The recent investment in AI and machine learning at the major tech companies has added a new demand source for talent that competes directly with what the auto industry needs for its own AI work.
The geographic competition has intensified. The same software-talent locations that the auto industry has been building presence in — Silicon Valley, Seattle, Austin — have become more competitive for software talent generally, with rising compensation, increased competition for senior engineers, and more turnover among the mid-career engineers who have historically been the auto industry's recruiting sweet spot.
The internal organizational dynamics
The auto industry's internal organizational dynamics around software engineering have been a persistent challenge that the industry has been working through with mixed success. The software engineering functions have historically been organized within engineering organizations whose principal orientation has been around mechanical and electrical engineering, with cultural and operational practices that have not always been well-suited to software development at the scale and pace the strategic priorities now require.
The compensation structures have been one of the most-visible friction points. The OEM compensation systems have historically been designed around the talent market for traditional automotive engineering, with compensation bands that align with the engineering-management hierarchy and that do not always accommodate the higher-compensation requirements of software engineers at competitive market rates. The major OEMs have generally adjusted their compensation systems to address the gap, but the adjustments have been incremental and have not always been sufficient to retain senior software engineers who can earn substantially more at consumer-tech competitors.
The career-path question has been another persistent friction point. The software engineers who join OEM organizations have historically faced career-progression structures that have emphasized management progression rather than technical-individual-contributor progression. The senior software engineer who wants to remain in a technical role rather than transition to management has not always found the OEM environment to support that career path as effectively as the consumer-tech environment does. The OEMs have been working to develop better technical-track career paths, but the transition has been gradual and uneven across organizations.
The operational practices for software development have been a third source of friction. The industry's historical engineering practices have been shaped by the requirements of mechanical and electrical engineering — long development cycles, careful design verification before manufacturing commitment, and the kind of waterfall-style program management that the safety-critical nature of vehicle development has supported. The software development practices that the consumer-tech industry has refined — agile development, continuous deployment, iterative product development — have been incorporated into the OEM software functions with varying degrees of success and have not always integrated cleanly with the broader vehicle-development practices the OEMs operate within.
What the partnerships are and are not delivering
The partnerships with external software companies have delivered substantial value for specific software capabilities that the OEMs could not have built independently. The infotainment platforms based on Android Automotive or on Apple's CarPlay are practical examples of partnership-delivered capability that has reached consumer-grade quality faster than OEM-internal development would have achieved.
The partnerships have limitations that are increasingly visible as the OEM software-strategy ambitions have grown. The partnership-delivered capabilities reside outside the OEM's own engineering control, with implications for product differentiation, for long-term strategic flexibility, and for the OEM's ability to evolve the capabilities at the pace and in the directions the product strategy requires. The OEMs that have built their software strategies most heavily around partnerships have found that the partnerships do not provide the strategic control over software direction that the OEMs increasingly want.
The OEMs have been working to recalibrate the partnership-versus-internal balance, with several major automakers having announced strategic shifts toward more internal software capability and reduced reliance on partner-delivered capabilities for the most strategically important software functions. The recalibration intensifies the talent-acquisition challenge because it shifts more of the software-development burden to internal organizations that are already constrained on talent.
What the next several years probably look like
The talent problem is going to get worse before it gets better. The combination of escalating talent requirements, more competitive talent markets, and internal organizational challenges means that the auto industry is going to be working through a multi-year period of software-talent constraint that is going to limit how quickly the software-defined-vehicle priorities can actually be delivered.
The successful OEMs will probably be the ones who address the organizational dynamics most effectively. The compensation systems, the career paths, and the operational practices that have constrained software-engineering effectiveness within the OEM environment can be improved, and the OEMs that improve them fastest will have advantages in talent acquisition and retention that compound over time.
The product-roadmap implications are real. The announced software-defined-vehicle timelines from most of the major OEMs are aggressive relative to what the talent constraints support, and the timeline slippage that has already started to become visible in the product-launch announcements will probably continue. The OEMs that deliver software capability at the pace their announcements suggest will probably be the exception rather than the rule, and the gap between announcement and delivery will be one of the more important things to watch in the industry over the next several years.
The software talent question is, at this point, one of the principal strategic constraints on the auto industry's transition to the software-defined vehicle model. The industry has the capital, has the manufacturing capability, has the regulatory positioning to deliver on the announced strategies. What it does not yet have in sufficient quantity is the software engineering talent at the scale and skill level the strategies require. The next several years are going to be substantially about whether the industry can solve that problem fast enough to deliver on the commitments it has already made.
