Tesla's FSD Count: A Metric Tailored to Executive Compensation

The most significant detail in Tesla's latest shareholder report is not the raw number of drivers using its autonomous software, but the specific methodology used to arrive at that number. The company claims to have reached 1.74 million active Full Self-Driving (FSD) subscriptions, a figure that has increased by 67% over the past year. However, this total is not a simple count of monthly recurring revenue. It is a composite metric that includes both customers paying a monthly fee and those who purchased the feature outright, a definition explicitly aligned with the terms of the CEO's 2025 performance award.

According to a footnote in the Q3 2026 shareholder letter, the metric follows the '2025 CEO Performance Award' and encompasses both upfront payments and monthly subscriptions. This distinction is critical because reaching 10 million active FSD subscriptions is a key operational milestone required for Elon Musk to earn his compensation package. By including owners who paid up to $15,000 for the software years ago, Tesla inflates the current engagement numbers without requiring new monthly cash flow from those specific users. This approach allows the company to report strong growth in a metric that directly ties to executive pay, even if the composition of that growth is mixed.

Tessla ceased selling FSD as a one-time purchase on February 14, 2026. Since then, new customers in the U.S. have been required to pay $99 per month to access the feature. Yet, every vehicle that purchased the software before that cutoff date remains in the active subscription count for as long as the car is operational. The only vehicles excluded from this total are those in free trial periods. This creates a scenario where a significant portion of the 1.74 million 'subscribers' are not currently contributing to monthly revenue, a nuance the company has not fully broken down in its public disclosures.

Company executives have previously stated that they believe FSD is driving demand in markets where it is available. In the third quarter, Tesla reported that more than 75% of new deliveries in North America included an FSD subscription. The company describes this as a nearly 20 percentage point increase from the previous quarter. While this attach rate suggests strong market adoption, Tesla has not defined whether this figure includes introductory offers or how many of these new buyers remain paying customers after the initial period. The lack of clarity on retention makes it difficult for investors to assess the long-term stickiness of the service.

The same accounting logic appears in Tesla's cumulative delivery figures, which now stand at 10.2 million. This number includes unsupervised robotaxis placed into commercial operation, again matching the criteria of the pay package which targets 20 million total vehicles. This alignment between reported metrics and compensation triggers raises questions about whether these numbers are designed for investor analysis or for satisfying the specific conditions of the CEO's contract.

Looking at the trajectory toward the 10 million milestone, the timeline appears stretched. Tesla added a net 260,000 FSD subscriptions in Q3, its best quarter to date. If the company maintains this pace, it would take approximately eight years to reach the 10 million target. If it relies on the average growth rate of the last 12 months, which is roughly 700,000 subscriptions per year, the timeline extends to nearly 12 years. The 2025 award runs until 2035, providing a long runway, but the current pace suggests the milestone is far off.

The ambiguity extends to the definition of 'active.' While free trials are excluded, the company does not specify how many of the 1.74 million users are actually paying the $99 monthly fee versus those who bought it outright. Tesla knows this breakdown, but it has not disclosed it. For a metric that serves as the headline for the company's software business, the lack of granularity leaves investors relying on a definition that was written for compensation purposes rather than business performance.

Ultimately, the issue is not that counting upfront buyers is incorrect; those vehicles do have the software. The concern lies in presenting this total as proof of momentum in a subscription-based model. With the board having written the definition and Musk's pay package depending on it, the metric becomes a dual-purpose tool. It reflects software penetration, but it also reflects the specific terms of the CEO's deal. Until Tesla provides a clearer breakdown of paying monthly subscribers versus legacy upfront buyers, the 1.74 million figure remains a complex number that requires careful interpretation.

The next logical step for transparency would be for Tesla to disclose retention rates for the new $99/month cohort. A 75% attach rate is impressive if those customers stay, but if many cancel after an introductory period, the true subscription base is much smaller. Until that data is public, the 1.74 million number is a best-case scenario for the pay package, not necessarily a definitive measure of the business's recurring revenue health.