Microsoft reported strong Q2 FY2024 earnings on January 30th with Azure growing 28% year over year. The AI contribution is increasingly quantifiable. The Microsoft-OpenAI relationship is paying off in ways that were not fully priced in when the investment was announced.
The Azure growth story
Azure's 28% growth included an estimated 6 points from AI services. That is a meaningful contribution from a product category that did not exist two years ago. The Azure OpenAI Service is the primary driver: enterprises paying for managed access to GPT-4 and other OpenAI models with Microsoft's data residency and compliance layer. The remaining 22 points come from the broader cloud migration and digital transformation spending that Azure had been growing on before the AI cycle.
The Microsoft 365 Copilot revenue picture
Microsoft added a $30/user/month surcharge for Microsoft 365 Copilot. At the Q2 2024 report, Copilot commercial seats were growing but still a small percentage of the 400 million Microsoft 365 commercial users. The penetration rate will be the story to watch over 2024 and 2025. Getting 10% of commercial Microsoft 365 users on the $30 Copilot tier would add $14.4 billion in annual recurring revenue. That is roughly the size of Salesforce's total annual revenue in 2020.
GitHub Copilot enterprise momentum
GitHub Copilot reached 1.3 million paying subscribers and GitHub announced Copilot Enterprise, a version with codebase-aware suggestions and pull request summarisation, for larger organisations. Developer tool revenue has historically been a small line in Microsoft's results. GitHub Copilot is the first developer tool at a price point and adoption rate that changes that.
The OpenAI dependency risk
Microsoft's AI revenue is substantially dependent on OpenAI models. If the OpenAI relationship deteriorated, or if a competing model substantially outperformed GPT-4 for enterprise use cases, Microsoft's AI revenue concentration would be a risk factor. Microsoft has been hedging: Azure OpenAI supports multiple models, including Llama 2 and Mistral. The custom Phi family of small language models that Microsoft Research is developing reduces dependency further. But the core commercial bet remains on GPT-4 class models from OpenAI for the foreseeable future.