Microsoft – On the Back Foot

Microsoft’s AI is mediocre, and there is no visible fix yet.

  • Satya Nadella’s tenure at Microsoft has been pretty faultless, but Microsoft’s problems with its AI offering are worsening, and the threat from Apple in the mid-tier of the market is very real indeed.
  • The net result is that I am becoming more sceptical with regard to its ability to grow as enterprises adopt AI, as outside of Azure, its in-house offerings are passable at best and are increasingly being seen as OS bloat rather than useful tools.
  • This is what has prompted Pavan Davuluri (EVP Windows and Devices) to lay out some sweeping changes that aim to address the criticism and keep users onside.
  • Microsoft has two ways to earn a return from the AI investments that it is making, one of which is to sell the compute for others (e.g., OpenAI), while the other is to offer AI services of its own to enhance the appeal of its existing products as well as sell services directly.
  • This is all well and good, but the problem is that its AI services are mediocre at best, meaning that no one really wants to use them.
  • For the last 2 years, Microsoft has been stating that AI was the future of Windows and that it was becoming an agentic OS, but this has not come to pass.
  • Instead, the most basic functions such as search within Office Apps and the file system remain execrable at best, recall simply does not work properly, and Copilot never seems to be able to do what is asked of it.
  • Consequently, Microsoft is pulling back from “infusing the OS with AI” to putting it in very specific places only and has also promised to fix other gripes about the taskbar and annoying updates.
  • At the same time, the MacBook Neo, which comes at $600 and represents surprisingly good value for money, as it does most things extremely well, represents a real threat in the mid-market.
  • Although users have much less freedom and choice when using a Mac, the vertical integration between software and hardware and minimal bloat in MacOS look more attractive as Microsoft fumbles around with AI in Windows
  • The reviews on the Neo are excellent, and I will not be surprised to see Apple start to eat away at Microsoft’s dominance at this price point.
  • The problem for Microsoft is that despite its investments in AI and early access to arguably the best AI technology available, its in-house products are awful, which means that no one wants them on their desktop or in their apps.
  • Hence, what began as a potentially early lead has pretty much been squandered, and Microsoft is currently losing the battle for both the consumer AI ecosystem to OpenAI and Google and the enterprise AI ecosystem to Anthropic.
  • However, in the business of providing compute to companies to train and run their AI, things continue to go very well, and it is not inconceivable that Microsoft will one day challenge AWS.
  • Microsoft must have grander plans than to be the engine room of AI, as its other businesses will require distinct features of their own if they are going to hold onto market share and margin in the business of offering software to companies.
  • This is where I see trouble because progress on this front is slow, and Microsoft is on no one’s lips when it comes to considering who the leaders of the consumer and enterprise AI ecosystems will be.
  • This is why Microsoft’s shares have already lost 28% of their value over the last 6 months, but I am not sure that this is enough.
  • The shares are currently trading on 22x June 2026 PER and 19x June 2027, which looks expensive when I can have Nvidia on 21x 2026 PER.
  • That being said, it is now cheaper than Meta, Google, Amazon and Apple, leading me to think that the market is already fairly well aware of this problem.
  • My concern is that Microsoft does not really have a fix for this in the pipeline, and so it could get cheaper yet.
  • However, it could equally just bumble along at the back of the pack and not unwind further, and so I remain completely indifferent to the shares here.

RICHARD WINDSOR

Richard is founder, owner of research company, Radio Free Mobile. He has 16 years of experience working in sell side equity research. During his 11 year tenure at Nomura Securities, he focused on the equity coverage of the Global Technology sector.

Leave a Comment