According to CNBC, Microsoft CEO Satya Nadella recently announced over 150 million people are using Copilot AI assistants across productivity, cybersecurity, and coding tasks. The company started selling the commercial version of 365 Copilot two years ago for $30 per person monthly as an add-on to its productivity suite. Microsoft’s Azure cloud business saw 40% revenue growth last quarter, outpacing Amazon and Google’s cloud divisions. However, conversations with IT buyers at Microsoft’s Ignite conference this week revealed significant resistance, with consultants reporting clients saying “I don’t even want it” about Copilot licenses despite the company’s $13 billion OpenAI investment and OpenAI’s separate $250 billion cloud commitment.
The enterprise AI reality check
Here’s the thing about enterprise software: companies will pay for infrastructure they need, but they’re ruthless about cutting tools that don’t deliver clear ROI. Microsoft‘s Azure growth shows businesses are absolutely investing in AI infrastructure – that’s the plumbing that runs everything. But Copilot is a different beast entirely. It’s a productivity tool that employees have to actually use and benefit from. And at $30 per person monthly? That adds up fast for a thousand-person company. Basically, you’re looking at $360,000 annually for something that might just become digital clutter.
The adoption challenge
Look, Microsoft has enterprise dominance locked down. Office 365 is everywhere. But that doesn’t automatically translate to AI chatbot adoption. The consultants CNBC spoke with highlighted this exact problem – companies are looking to reduce their Copilot licenses, not expand them. Why? Because the return isn’t clear yet. When you’re buying cloud infrastructure, you know exactly what you’re getting. But an AI assistant? It’s much harder to measure whether it’s actually making employees more productive or just giving them a fancy toy. And in business technology, if you can’t measure the benefit, you can’t justify the cost.
Microsoft’s AI gamble
Microsoft is betting big on multiple AI fronts simultaneously. There’s the infrastructure play with Azure, the partnership with OpenAI, and now the productivity tools with Copilot. The infrastructure part seems to be working beautifully – 40% growth in a competitive cloud market is impressive. But the tool adoption? That’s the uphill climb. It reminds me of other enterprise software struggles where companies buy licenses that just sit unused. The real test will be whether Microsoft can demonstrate tangible business outcomes that justify that $30 per seat. Otherwise, we might see more enterprises doing exactly what Adam Mansfield described – wanting those license counts to go from 300 to zero.
