Tesla Promised The Future, Waymo May Be Delivering It

Tesla (NASDAQ:TSLA) has long positioned itself as a leader in autonomous driving. Its Full Self-Driving (FSD) software enhances the driving experience by automating certain tasks, allowing Tesla vehicles to navigate, change lanes, and park themselves while requiring the driver’s active supervision. Yet, despite years of hype and lofty investor expectations, Tesla’s FSD doesn’t appear to contribute significantly to Tesla’s revenues, while other tech titans, such as Google, have been increasing their investments in self-driving technology. At the same time, higher-margin software products like FSD are becoming increasingly important to Tesla’s growth narrative as its core EV business comes under pressure from softening demand and growing price competition. So, where does the FSD software stand at the moment?

$550 Million in FSD Revenue

Tesla offers FSD as either an $8,000 upfront purchase or a $99-per-month subscription. We believe that the subscription model is likely to be the go-to option for most new Tesla customers, given its flexibility and lower upfront cost. Although Tesla doesn’t break out FSD revenue separately, estimates from RBC Capital Markets suggest that about 5% of customers had opted for the subscription as of early 2024. The attach rate is very likely to have improved after Tesla slashed the monthly price by 50% to $100 by April 2024. Based on vehicle delivery trends for 2025, we estimate that Tesla could sell about 1.7 million vehicles this year. If we assume that 10% of these customers opt for the FSD subscription, that adds up to about 170k new subscriptions.

That would bring cumulative subscriptions to about 460k by the end of this year. See detailed calculations in our dashboard How Big Is Tesla’s Software Business? At a monthly rate of $100, that’s about $550 million in annual revenue for Tesla. That’s meaningful, but still a fraction of Tesla’s total sales, and far from justifying the massive valuation premium Tesla stock gets based on FSD’s long-term potential. Separately, if you seek upside with lower volatility than individual stocks, the Trefis High Quality portfolio presents an alternative, having outperformed the S&P 500 and generated returns exceeding 91% since its inception.

Is Waymo Ahead Of Tesla?

Elon Musk has repeatedly promised major advances in autonomy, but Tesla has often missed its own deadlines. For instance, Musk previously claimed that v13 of the FSD software would enable a 5 to 6x increase in miles between disengagements compared to the v12.5 version.  The miles between critical disengagements, which is the average number of miles a vehicle drives before the system must be disengaged, is seen as a key safety metric for FSD. However, data now shows that v13 barely brought a 2.5x improvement, going from 200 miles to about 495 miles. The head of FSD at Tesla has previously stated that for Tesla to enable truly unsupervised self-driving, Tesla would need to achieve the average in miles per critical intervention “equivalent of human miles between collision,” which stands at about 700,000 miles. Tesla appears to have a long way to go by this yardstick.

On the other hand, Alphabet’s Waymo has been quietly and consistently making real-world progress. The company is now operating over 250,000 fully autonomous, paid rides per week across Phoenix, San Francisco, and Los Angeles, up from around 10,000 rides just two years ago. Waymo’s vehicles drive without any human supervision, unlike Tesla’s FSD, which still requires driver oversight. Riders, too, seem to love Waymo. Per Earnest Analytics, Waymo retains riders at a higher rate than Uber or Lyft. This suggests that once people try fully autonomous rides, they prefer them. Safety could be another selling point. Waymo’s safety report last year noted that its autonomous vehicles achieved a 78% reduction in injury-causing crashes compared to human drivers. See: Waymo To Separate From Google?

Tesla May Have Tricks Up Its Sleeve

That said, Tesla has some advantages at this point. Waymo’s system is built around a comprehensive hardware stack including custom lidar and radar sensors as well as cameras that are retrofitted to the cars it buys from mainstream automakers. Tesla, in contrast, bets on a camera-only approach powered by neural networks. This makes Tesla’s system potentially easier and cheaper to scale. Moreover, there are likely over 5 million Teslas on the road today, and it’s safe to assume that a large percentage of these vehicles have the necessary hardware to run FSD software. If the software can catch up, Tesla could flip the switch on a massive FSD user base, quickly transforming the economics of its business.

Invest with Trefis Market-Beating Portfolios

See all Trefis Price Estimates