Table of Contents
Imagine hailing a car on a rainy San Francisco day, but when it arrives, there’s no one in the driver's seat. The vehicle navigates traffic, inches into a busy lane, and safely delivers you to your destination, all on its own. This isn't science fiction; it's the daily reality of Waymo's autonomous ride-hailing service. We sat down with Shweta Shrivastava, Senior Director of Product Management at Waymo, to peek under the hood of what it takes to build a product that is redefining mobility. With a background spanning Amazon, Cisco, and AI startups, Shweta shares invaluable lessons on product strategy, team leadership, and the unique challenges of building trust between humans and intelligent machines.
Key Takeaways
- Trust is Engineered Through Nuance: Waymo builds rider confidence not just with flawless safety records, but by programming its vehicles with subtle, human-like behaviors, like slowing down on steep hills or cautiously inching into traffic to signal intent.
- Product Management for a Driverless World: PMs at Waymo must possess deep technical knowledge and embrace a long-term vision. The concept of a Minimum Viable Product (MVP) is redefined, as the bar for safety is non-negotiable from day one.
- Measuring Success in Autonomy: Progress is tracked through a dual lens of commercial metrics (trips per week, active users) and sophisticated system performance metrics, ensuring the Waymo Driver is safer and more reliable than a human driver.
- Disrupt Yourself Before You're Disrupted: A key lesson from large tech companies is the need for proactive self-disruption. Product leaders must constantly challenge the status quo to avoid being outmaneuvered by nimble startups.
- Listening is a Superpower: Shweta emphasizes that foundational PM skills like listening and empathy are crucial for influencing without authority. The best product leaders challenge their own assumptions and seek out contention to uncover the truth.
The Human Touch in Autonomous Driving
One of the most fascinating challenges for Waymo is teaching a machine to navigate a world governed by unwritten social rules. Driving isn't just about following traffic laws; it's a complex, interactive, and social dance. Shweta explains that creating a comfortable and trustworthy experience means making the vehicle's behavior feel natural and predictable, not robotic.
Building Trust Through Body Language
When a Waymo vehicle needs to merge into a busy lane, it can't make eye contact with other drivers. Instead, it uses its "body language"—subtly inching forward to signal its intention. This behavior is a direct result of training deep learning models on vast amounts of human driving data. The system learns to mimic good human driving habits, creating a sense of familiarity for both riders and other cars on the road.
We're using a lot of human driving data to train our deep learn models so it's important to make sure that the behavioral car doesn't seem robotic.
Adapting to Social Norms
Waymo's system goes beyond explicit gestures. It learns to understand and adapt to local social norms. For example, in some parts of San Francisco, pedestrians might cross the street even without a walk signal. The vehicle must learn to anticipate and react to these location-specific behaviors. This adaptability is key to making the service feel seamless and intelligent.
Designing for Predictability
To build trust from the moment a rider steps inside, every element of the experience is designed to be credible and predictable. Riders can see what the car sees on an in-car monitor, and a support team is just a call away. Interestingly, Shweta shared an example of how rider feedback refined the car's behavior. While an autonomous vehicle could safely maintain the speed limit downhill, Waymo learned that human drivers instinctively slow down. The team modified the behavior to match this expectation, making the ride feel more natural and trustworthy.
Product Management in the World of Atoms and Bits
Building a fully autonomous vehicle is a fundamentally different challenge than creating a purely software product. Shweta, who has led product teams at AWS and Cisco, highlights the unique demands placed on product managers at Waymo, where software meets the physical world with life-or-death stakes.
A Higher Bar for MVP
In the world of SaaS, the "Minimum Viable Product" is a core concept used to ship quickly and iterate based on user feedback. At Waymo, this philosophy takes on a new meaning.
The concept of MVP...has a whole new meaning here at Waymo... The MVP bar itself for safety is extremely high for us.
You cannot cut corners on safety. While the principle of iteration still applies, the initial product must meet an incredibly high standard of safety and reliability before it ever reaches a public rider. This requires a different mindset, focusing on rigorous validation and simulation before deployment.
Technical Depth and a Long-Term Mindset
Product managers at Waymo must be able to go deeper technically than in many traditional software roles. They need to understand the complexities of AI, simulation, and onboard systems to make informed decisions. Furthermore, solving autonomous driving is a long game. It requires tenacity and a mission-driven focus on making roads safer. As Shweta notes, the mission to reduce the 1.35 million annual traffic deaths worldwide is a powerful motivator for the team.
Measuring Progress on the Road to Autonomy
How do you measure progress on a project as ambitious as a fully autonomous vehicle? It's not as simple as tracking miles driven. Waymo employs a sophisticated framework of metrics that cover everything from business growth to the nuanced performance of the Waymo Driver.
Commercial and Operational Metrics
Waymo is not a research project; it's a live, paid commercial service in cities like Phoenix and San Francisco. As such, the product teams track standard commercial metrics you’d see at any tech company, including:
- Trips per week
- Daily and weekly active users
- Customer acquisition funnel metrics
- Operational costs per trip
These KPIs ensure the team is focused on building a sustainable business, not just groundbreaking technology.
System Behavior and Safety Metrics
The core of Waymo’s progress is measured by the performance of the "Driver" itself. The ultimate goal is to be significantly safer than a human. This involves extensive data collection to establish a benchmark for human driver performance (e.g., collisions per million miles) and then rigorously measuring Waymo's system against it. Other key performance metrics include:
- Progress: Ensuring the vehicle gets to its destination on time without getting unduly stopped or stranded in traffic.
- Compliance: Adhering to all road rules and regulations.
- Comfort: Providing a smooth and natural-feeling ride for passengers.
Lessons from a Career Spanning Tech Giants and Startups
Across her career, Shweta has accumulated a wealth of wisdom about building successful products and teams. Whether at a massive corporation like Amazon or a cutting-edge company like Waymo, she believes certain core tenets of product management remain universal.
Work Backwards from the Customer
A practice famously championed by Amazon is to write a press release for a product before a single line of code is written. This forces the product manager to clearly articulate the customer problem and the product's value proposition. Shweta notes that while Waymo has its own version of this process, the core principle is the same: always start with who you're building for and what problem you're solving.
The Power of Saying No
Just as important as knowing what you're building is knowing what you're not building. It's easy to get distracted by various customer requests, but a product that tries to be everything to everyone often fails. Maintaining a sharp focus and being crisp about prioritization is critical for success, both at startups and large companies.
Disrupt Yourself Before Someone Else Does
You need to disrupt yourself before somebody else does.
This is the classic innovator's dilemma. Large, successful companies can become complacent and focus only on incremental improvements, leaving them vulnerable to upstarts. Shweta champions the mindset of a product leader who constantly challenges their own assumptions and business models. She was part of the team at Amazon that launched Honeycode, a no-code application platform, which was a move into a brand-new space for the company—a prime example of self-disruption.
Developing the Core Skills of a Product Leader
Beyond frameworks and processes, what truly separates great product leaders are their interpersonal skills and mindset. Shweta points to a few underrated qualities that she believes are essential for career growth and impact.
The Underrated Skills: Listening and Empathy
While skills like prioritization and writing specs are often highlighted, Shweta argues that listening and empathy are the true foundation of influence. To lead without authority, a PM must genuinely understand the constraints and motivations of their engineering, design, and business counterparts. This requires coming in with a beginner's mindset, ready to absorb information before jumping to conclusions.
Challenge Your Own Assumptions
Effective listening isn't passive. It's an active process of questioning your own beliefs. Shweta advises PMs at all levels to constantly ask themselves, "Am I challenging my own assumptions?" If there's no conflict or contention in discussions, something is likely being missed. Healthy debate is a sign that a team is rigorously pursuing the best solution.
How to Get Promoted
When asked about getting promoted, Shweta offers a counterintuitive insight: the best way to get promoted is to not want it too badly. Instead of optimizing for a title change, focus on delivering impact for the business. Make your ambitions known to your manager, but dedicate your energy to working on the right things and solving the most important problems. If you consistently create value, recognition and advancement will naturally follow.
Conclusion: The Road Ahead
The journey to a fully autonomous future is a marathon, not a sprint. It requires a unique blend of technological innovation, deep human empathy, and disciplined product development. Shweta Shrivastava's insights from the front lines at Waymo reveal that success isn't just about building a car that can drive itself—it's about building a service that people can trust, enjoy, and integrate seamlessly into their lives. For product leaders everywhere, the lessons are clear: stay focused on the customer, don't be afraid to disrupt yourself, and never underestimate the power of listening. If you're in Phoenix or San Francisco, you can experience this future today by downloading the Waymo app. The driverless car is no longer a distant dream; it's here, and it's waiting for its next rider.
){kind=link}