In the wave of automotive intelligence transformation, traditional OEMs are diving into software—establishing software subsidiaries to develop autonomous driving and smart cockpit features. But these attempts face enormous risks.
Why Do OEMs Build Software Companies?
1. Out of Greed
In February 2020, Tesla's market cap soared to #1 among global automakers. In October 2021, it broke trillion. Tesla's success showed traditional OEMs the massive market potential. Musk repeatedly emphasized Tesla is not just a car company—it's an autonomous driving company, an AI company.
This valuation gap made traditional OEMs panic—they thought if they just did autonomous driving and full-stack in-house development like Tesla, they'd get the same market premium. But they ignored their own strengths and weaknesses. If you're a casino owner who makes the rules, you always win long-term. Then some kid tells you to play by his rules—now you're on equal footing.
2. Out of Fear
Beyond greed, another motive is fear.
"If traditional OEMs don't transform, they become contract manufacturers for new EV makers."
"If traditional OEMs don't transform, they become the next Nokia."
These warnings aren't baseless. The smartphone industry showed that giants can collapse quickly against new challengers. OEMs know this threat is real, but the electric+smart wave is upon them and they don't know how to stop it—hence fear.
5 OEM Strategies
Strategy 1: Internal Small Team
Approach: pull together a small internal team, hire a few external experts, and get going.
Pros: minimal trial-and-error cost, fast start.
Cons: extremely difficult—no successful precedents. Autonomous driving and smart cockpit require deep technical reserves that traditional OEMs lack. External experts struggle to fit into traditional OEM organizational culture.
Strategy 2: Independent Software Subsidiary
Examples: SAIC Zero-Base, Chery Dazz, VW CARIAD.
Pros: independent decision-making, hiring, funding, and order flow—every founder's dream start.
Cons: although middle management is externally hired, executives are parent-company appointees who understand traditional OEM decision-making: "do less, make fewer mistakes." All decisions require layers of approval. If sales underperform for any reason, the software subsidiary bears the consequences. Cut off from parent funding, they can't survive independently.
Strategy 3: Joint Venture with Software Company
Examples: Banma, Qiruicheng.
Pros: risk-sharing, each side gets what they need.
Cons: obvious interest conflicts. OEMs want advanced technology exclusively for themselves; suppliers want their tech reused across brands. One side has deep OEM DNA—requiring full visibility into execution for quality control; the other is an emerging software supplier prioritizing efficient development with minimal documentation.
Strategy 4: Buy Solutions Directly
Examples: Seres buys Huawei solution, VW VCTC buys XPeng solution.
Pros: fast results—quickly level up intelligence. Seres became profitable by betting on Huawei.
Cons: long-term risk of "teaching the apprentice, starving the master." VW understood this—before seeking XPeng's help, it invested in XPeng as a peace offering. VW will learn XPeng's team management, architecture design, iteration development, and infrastructure toolchain knowhow—enough to gain supplier negotiation leverage, not necessarily to build everything in-house.
Strategy 5: Scatter Investments Downstream
Examples: investing in Momenta, Horizon Robotics, etc. OEMs act as investor and customer, without seeking control.
Pros: minimal risk and investment, potentially high long-term returns. Maintains healthy supplier relationships while flexibly adjusting investment over time.
Cons: requires strong market judgment and strategic discipline from the OEM.
Summary
A clear pattern emerges: the more hands-off the OEM, the smoother the collaboration with software suppliers; the more the OEM rolls up its sleeves to do the work itself, the more mud it gets on its clothes.
Traditional OEMs shouldn't compare software development speed with new EV makers—that's adopting someone else's rules and competing on their strengths.
