Recently, the Chief Technology Officer (CIO) of a large pharmaceutical company canceled a contract Artificial Intelligence (AI) with Microsoft after just six months of use. The decision to cancel the implementation of Office 365 Copilot, a tool that promises to improve productivity through AI, raises significant concerns for the technology industry as a whole.
The Reason for Cancellation
According to a transcript of a call with Morgan Stanley analysts, published in a Business Insider article, the CIO, identified only as Greg, explained that the company had invested significantly in enabling 500 employees to use Office 365 Copilot in the fourth quarter of 2023 and the first quarter of 2024. However, after six months of use, the company concluded that the AI tools offered by Microsoft did not justify the additional cost.
Greg pointed out that Microsoft's AI slide-generating functionality was comparable to "high school presentations," emphasizing that the product's quality didn't match the price. The E3 version of Microsoft 365, which costs about $34 per user per month, had an additional $30 per user per month to include Copilot's features. For 500 employees, this extra cost represented approximately $180.000 per year, an amount that, according to Greg, wasn't justified given the benefits offered.
Legal Issues and Marginal Utility
While Microsoft Teams' video meeting archiving and summarization feature was the most promising, the pharmaceutical company's legal team expressed concerns about the retention of meeting transcripts, leading to the decision to discontinue this functionality. This left the company with tools that Greg described as "marginally useful," comparing the use of AI in PowerPoint and Excel to basic functions that didn't justify the additional investment.
Implications for Microsoft and the AI Market
This case highlights a critical issue for Microsoft and the entire technology industry: the ability to create real value with AI solutions. Microsoft has invested heavily in AI, amassing 1,8 million graphics processing units (GPUs) to build and run its AI models, and plans to triple the capacity of its data centers, primarily to handle AI workloads. These massive investments generated a record $14 billion in capital expenditures last quarter.
However, Greg's comments raise questions about whether these investments will be enough to convince enterprise customers to pay more for these solutions. This is particularly concerning considering that Microsoft is considering revamping how it licenses its tools, possibly bundling AI functionality into existing packages or creating a more expensive offering with Copilot capabilities.
My Personal Vision
As I've mentioned in other posts, when it comes to AI, many companies have focused on tools without fully understanding the problems they want to solve. Microsoft's Copilot is a clear example. Costing $30 per user per month, it promises, for example, to automatically create PowerPoint presentations. But, as the CIO in question noted, the quality of these presentations is questionable. And even if they were good, the question we should ask ourselves is: "How many PowerPoint presentations does a typical user create per month?"
Investing in an expensive tool to solve a problem that doesn't actually exist is unlikely to add value. It's crucial that companies are clear about the problems that need to be solved before adopting a new technology. Otherwise, we run the risk of wasting resources on solutions that, in practice, don't add value to the business.
Reflection for Industry
This major pharmaceutical company's decision to cancel its AI contract with Microsoft serves as a warning to the entire tech industry. If a company of this size, with the resources to invest in cutting-edge technology, fails to see the value in AI tools costing hundreds of thousands of dollars, other companies may follow suit.
This highlights the need for technology vendors to focus on creating solutions that deliver tangible and justifiable value, especially in an economic environment where return on investment is increasingly scrutinized. Microsoft, in particular, needs to reevaluate how its AI offerings are perceived and used by its customers if it wants to maintain its market leadership.
This case is a clear example that, however promising emerging technologies may seem, they need to deliver concrete and measurable results to ensure their adoption and long-term success.
