PLTR's new vision for TAM

Yesterday, Goldman Sachs released a research report with the core view that PLTR's TAM is poised for leapfrog growth. Why do they say this? I attempt to break it down from three perspectives:

PLTR is no longer just a 'government data platform + high-growth software' but is entering a second-layer TAM market: budgets paid by institutions for Sovereign AI and industry-specific production applications, which are deeper and broader than traditional data analytics solutions.

Three drivers, each 'expanding the TAM':

1. Sovereign AI
Old TAM boundary: Large enterprises buy BI/data/integration, with budgets categorized under IT sub-items.
New TAM: Governments, defense, critical infrastructure, and regulated industries create standalone projects to keep data, weights, and learning loops within their institutions—such as ZDR, Reference Architecture, and anti-token colonialism.

Expanding from software subscriptions to AI transformation, national security, and compliance infrastructure results in higher single deal ACV and stickiness. Goldman believes this isn't PLTR eating into an existing pie, but rather a new pie growing.

2. Bespoke Applications
Old understanding: Palantir = generic platform, where customers build their own apps.
New understanding: Delivering runnable industry workflows directly on AIP, shifting from 'platform licensing' to 'platform + application layer value'.

For customers: They are buying business outcomes (supply chain, maintenance, compliance, combat coordination, etc.), not empty ontologies.
For TAM: The same customer can expand to more workloads and departments; the platform shifts from 'one project' to a multi-app host—deepening engagement.

3. Verticalization
Old path: Selling Foundry horizontally, requiring a fresh pitch for every client.
New path: Packaging by industry (solutions, templates, compliance scripts, reference architectures), shortening procurement decision-making and implementation cycles.

TAM Mechanism: Breadth: Mid-sized clients can enter using 'industry packages'
Replicability: Once one vertical is penetrated → marginal costs for peers in the same industry drop → what Goldman calls 'another set of higher estimate revisions.'

Goldman believes PLTR's TAM will grow 'leapfrog-style' because budget categories have expanded from 'data tools' to 'sovereign AI infrastructure + industry production applications.' In other words, Sovereign AI means clients are willing to pay more for compliance and security costs to 摆脱 being bottlenecked by token/model dependencies. This represents a new market space for PLTR. Meanwhile, AI FDE prevents this expansion from being linearly constrained by headcount. It's quite an interesting take.

$Palantir Tech(PLTR.US)