Samsara Inc. (IoT): Investor Day Takeaways: Advancing the Physical AI Opportunity

Research evidence excerpt

Samsara Inc. (IoT): Investor Day Takeaways: Advancing the Physical AI Opportunity

Goldman Sachs Samsara Inc. (IOT)

drew was instructive: a single customer’s fleet might offer roughly 1,000 vehicles and 20

of the same make, model and year, against Samsara’s ecosystem of millions of vehicles

and 10K+ of the same configuration, sitting on 60T+ proprietary data points and 99% of

US road coverage. The advantage compounds – annual data capture has risen from 2T

(FY21) to 25T+ (FY26) – and each connected asset sharpens the model and widens a

lead competitors cannot easily replicate. We think the payoff shows up in what that scale

now produces: agents and detections more accurate than any single customer or

horizontal model could replicate, and a steady cadence of monetizable AI products on

the same installed base. We see this as what makes the AI beneficiary case real, and it

moves the conversation past whether Samsara is merely a data source to how much of

the AI opportunity it can capture from a position of control.

Agents mark a shift in how Samsara monetizes – from connecting assets to

automating work. The three launch agents – Maintenance, Safety and Dispatch – each

automate an operational role where labor is the dominant cost and the work is

repetitive, and each is billed on usage rather than per asset. That ties revenue to value

delivered: Maintenance, demoed live from fault code to work order to warranty check to

vendor match, was pitched against a 10x-plus ROI on the labor it replaces. Two features

make the model attractive to us. The revenue is purely incremental – agents expand an

account rather than recut it – so adoption should flow through to net retention as usage

scales. And agents sell into deployments customers already run, where the hard change