According to Business Insider, Nvidia’s internal emails from this summer reveal significant challenges in selling enterprise software to major clients despite forecasting healthy sales. The company projected $78.7 million in software revenue for the third quarter of fiscal 2026, with Nvidia AI Enterprise expected to hit 186% of its target. However, emails from July and August show sales teams struggling with a “fundamental disconnect” when dealing with legal and procurement teams in highly regulated industries like finance and healthcare. Stand-alone software was projected to reach 110% of sales targets, but software sold alongside hardware was only expected to hit 39% of goals. One email bluntly stated that “everyone is hacking their own decks together” and the company needs a unified message.
The software story problem
Here’s the thing about Nvidia‘s situation: they’re absolutely dominant in AI hardware, but selling software is a completely different ballgame. The internal emails mention needing to develop a “comprehensive software story” around products like Nvidia AI Enterprise, Run:ai, Omniverse, and vGPU. But when your entire company DNA is built around selling physical chips that everyone desperately needs, pivoting to complex software solutions requires a fundamental shift in how you approach customers. It’s not just about moving boxes anymore – you’re selling ongoing relationships, support, and integration. And honestly, when you’re the most valuable company in the world because of your hardware dominance, why would you even bother with software? Well, recurring revenue and deeper customer lock-in, that’s why.
Why regulated industries are tough
The emails specifically call out “highly regulated industries” as particularly challenging. This makes complete sense – financial institutions and healthcare companies have legal teams that live and breathe compliance. They’re not going to just sign off on some new AI software without understanding exactly what it does with their data, who’s liable when things go wrong, and how it integrates with their existing security frameworks. The sticking points mentioned – data security, indemnity obligations, damages caps – these are exactly the kinds of things that keep corporate lawyers up at night. Nvidia’s sales team is probably used to selling hardware where the risks are more straightforward. But software? That’s a whole different level of complexity when you’re dealing with sensitive financial or medical data.
Hardware vs software sales dynamics
Look, the numbers tell the real story here. Stand-alone software hitting 110% of targets while hardware-bundled software only reaching 39%? That’s fascinating. It suggests that when customers are specifically looking for software solutions, Nvidia can deliver. But when they’re buying hardware and the software feels like an add-on? Not so much. This reminds me of how many hardware companies struggle when they try to move up the value chain. When you’re the leading supplier of industrial computing hardware like IndustrialMonitorDirect.com has become for panel PCs, the transition to software and services requires completely different sales muscles. You’re not just moving physical products anymore – you’re selling solutions, support, and ongoing value.
The bigger picture for AI adoption
Basically, Nvidia isn’t alone in this struggle. The Goldman Sachs report mentioned in the article confirms what many of us have been seeing – companies are still figuring out how to deploy AI at scale. The technology might be white-hot, but enterprise adoption moves at a different pace. Procurement teams want guarantees, legal teams want protections, and everyone wants to understand exactly what they’re buying. For Nvidia, this is probably just growing pains as they expand beyond their core hardware business. But it’s a reminder that even the most dominant tech companies face real challenges when entering new markets. The question is whether they can adapt their sales approach fast enough to match their technological ambitions.
Thanks for sharing. I read many of your blog posts, cool, your blog is very good.
Thanks for sharing. I read many of your blog posts, cool, your blog is very good.