CompanyGraph
NetApp, Inc.
NTAP · United States
Runs its storage software natively inside Amazon, Microsoft, and Google's clouds so that data behaves the same whether it sits in a customer's own data center or any of the three clouds.
NetApp runs a storage operating system called ONTAP that executes natively inside Amazon, Microsoft, and Google's cloud infrastructure — not as software sitting on top of it, but as a kernel-level service that each hyperscaler had to co-develop deep integration agreements to enable. Because the same ONTAP code runs both inside those three clouds and on a customer's own hardware, data can move between a hospital's on-premises array and any of the three cloud environments without a translation layer, which is what allows organizations with strict compliance rules to replicate data across borders without rebuilding their architecture. Customers who want to leave face months of migration work because SnapMirror replication relationships, Cloud Insights monitoring hooks, and StorageGRID namespace structures are all built against ONTAP's APIs and cannot be straightforwardly moved to a different storage platform. The whole structure depends on all three hyperscalers keeping their integration agreements intact — if Amazon, Microsoft, or Google deprecates its native ONTAP service or steers customers toward its own competing storage product, the unified management layer breaks for every customer whose setup crosses that provider.
How does this company make money?
NetApp earns money in four ways. It sells physical hardware — All-Flash FAS storage arrays and StorageGRID appliances — to customers who run storage in their own buildings. It sells ONTAP software licenses separately for on-premises deployments. It charges based on how much storage capacity a customer actually uses on Azure NetApp Files and Amazon FSx. And it collects subscription fees from customers who use Cloud Insights for monitoring and Spot for cloud cost optimization.
What makes this company hard to replace?
Customers who want to leave face three concrete problems. First, SnapMirror replication relationships are woven through their entire hybrid infrastructure, and migrating that data to a different storage platform takes months. Second, Cloud Insights monitoring is connected to customers' DevOps tools through APIs that would need to be rebuilt from scratch for any alternative. Third, StorageGRID object storage namespace structures cannot be directly exported to S3-compatible alternatives, so moving away from StorageGRID means reorganizing how the data itself is stored and addressed.
What limits this company?
AWS, Azure, and Google each control the computing and networking layers that sit beneath ONTAP's code, so if any one of them changes its infrastructure, NetApp's engineers have to update that cloud's version of ONTAP to keep it working the same way as the others. NetApp must run three separate cloud development tracks at the same time as it updates the on-premises version, and how fast that can happen is set by each cloud provider's own schedule — not NetApp's.
What does this company depend on?
NetApp cannot operate without five things: the ONTAP operating system itself, which NetApp develops and maintains as the core of everything; Amazon Web Services infrastructure, which hosts FSx for NetApp ONTAP; Microsoft Azure infrastructure, which hosts Azure NetApp Files; Google Cloud Platform, which hosts Google Cloud NetApp Volumes; and NVIDIA's hardware and software stack, which underpins the AI Data Engine platform.
Who depends on this company?
Healthcare organizations running electronic health records depend on NetApp to move patient data between their own data centers and cloud environments in a way that stays compliant with HIPAA — if NetApp stopped, that compliant data mobility would break. Financial services firms use NetApp to replicate trading data in real time across hybrid infrastructure, and losing that would disrupt systems that cannot tolerate delays. Operators running large AI workloads depend on NetApp to manage exabyte-scale storage for the datasets used to train AI models.
How does this company scale?
ONTAP software licenses and cloud service subscriptions can be sold to new customers in new places without building anything physical — that part scales cheaply. What does not scale easily is the engineering work required to keep the on-premises version of ONTAP and all three cloud versions behaving identically, because each cloud provider changes its infrastructure on its own timeline and each change requires dedicated development effort from NetApp.
What external forces can significantly affect this company?
Data sovereignty laws in various countries require that certain data stay within specific geographic borders, which forces customers to split their storage across jurisdictions and complicates any architecture that spans multiple regions. Federal Reserve stress testing rules push financial services firms toward hybrid cloud setups to prove they can keep operating under pressure, which shapes how those customers buy and use storage. GDPR's right-to-erasure rules require that data be deletable across every storage tier — on-premises arrays, cloud volumes, and object storage — which adds compliance requirements to every layer NetApp manages.
Where is this company structurally vulnerable?
If Amazon, Microsoft, or Google decided to shut down its native ONTAP integration — by ending the agreement, blocking the kernel-level access ONTAP needs, or pushing customers toward its own competing storage product — the data management API would stop working the same way across that provider's environment. Every SnapMirror replication path that touches that cloud would fail, and any customer whose hybrid storage setup spans that provider would lose the unified management they built their architecture around.