Don't Be Microsoft

I want to be clear about something before we get into this: I use Microsoft products. I have for years. Outlook, Teams, Excel. They're genuinely good tools, and I'm not writing this to pile on a company that, by any financial measure, is doing fine. Revenue for its 2026 fiscal year, which ended in June, was $331.8 billion, up 18%. (source) Fine doesn't quite cover it.

But Microsoft just made a move that I think is worth paying attention to. Not because of what it does to their stock price, but because of what it signals. If you run a small business, the signal matters more than the number.

The move

On October 26, Microsoft offered voluntary buyouts to about 60,000 employees. Fortune puts that at about 27% of the company's workforce, more than a quarter. (source, source) The offer covers the United States, Canada, India, and the United Kingdom, and it's open to employees at the vice president level and below on select teams. (source) Microsoft calls it a "Voluntary Exit Offer." The memo from its chief people officer said the offer "gives eligible employees who may be considering a future outside the company a new choice."

This wasn't Microsoft's first buyout this year. In April, it offered a voluntary retirement buyout to about 8,750 US employees. That was the first program of its kind in the company's history. (source) Bloomberg reported that more than 30% of the people who were eligible accepted. (source) Now the offer is about seven times bigger.

Microsoft hasn't said the buyouts are about AI spending. On its October 28 earnings call, its chief financial officer said only that the company's cost guidance includes an estimate for the "recently announced employee voluntary exit offer." (source) But look at the context. Microsoft spent $45.2 billion on capital expenditures last quarter, and it says it expects to spend over $50 billion next quarter. (source) In its 2026 fiscal year, it brought 88 new data centres online. (source)

My read is that Microsoft is trimming its people costs while it pours money into infrastructure. I can't prove that, and Microsoft hasn't said it. But it looks like a very large, very public bet that the future belongs to the machine, not to the people running it.

I don't think that bet is the right one. And I think the early evidence is, at best, mixed for them.

The gap between the bet and the results

Here's the number that stopped me when I was researching this post. In July, Microsoft said Microsoft 365 Copilot, their flagship AI product, had over 30 million paid seats. That was double the 15 million they reported in January, so it's growing fast. (source, source) On the October 28 call, Microsoft didn't update that number. It said daily Copilot users were up nearly 50% from the previous quarter. (source)

That's real growth, and it deserves to be said plainly.

But Microsoft has around 450 million paid commercial seats of Microsoft 365. If the total is now about 464 million, as one report estimates, then 30 million Copilot seats is roughly 6 or 7%. (source) That means more than 90% of their own commercial seats still aren't paying for it. After 88 new data centres in a single fiscal year. After the buyouts.

Wall Street liked the quarter. Azure grew 48%, and the stock rose about 4% after hours. (source) So the market isn't punishing the bet right now. But the gap between the scale of the investment and the scale of the uptake is still large. And the question I keep coming back to is this: at what point does the machine have to prove it was worth more than the people?

That question doesn't have an answer yet. But it's the right one to be asking.

People are paying attention, and making choices

Here's what I'm seeing, and I'll be honest that some of this is observation rather than hard data: people are quietly making different choices about the tools they use, and privacy seems to be part of why.

According to StatCounter, Windows fell to 56.55% of global desktop usage in June 2026, below 60% for the first time in years. Linux reached 4.39%. (source) StatCounter counts visits to websites, not computers that are actually installed, so it's a rough guide. But the direction is worth noticing.

Some of that is the hardware requirements for Windows 11. Some of it is gaming on Linux getting better. And the coverage of those numbers points to frustration with how much data Windows collects, how hard Microsoft pushes its online services, and a sense that the product is increasingly designed around Microsoft's interests rather than the user's.

That's a slow leak. But slow leaks sink ships.

And it goes beyond software preferences. In my experience, customers today pay close attention to how companies treat their employees. They notice when a headline says a company is offering buyouts to more than a quarter of its workforce while it spends billions on AI. They make mental notes. They talk about it. And they factor it into decisions, sometimes consciously, often just as a quiet pull that shifts where they spend.

We've written before about how your customers can feel how you treat your team. Microsoft is demonstrating something right now. I expect the market to respond slowly, in ways that won't show up in a quarterly report for years.

No company lasts forever

The East India Company was, for a time, one of the most powerful commercial enterprises in the world. It had its own army, 260,000 strong at its height, twice the size of Britain's standing army. It had its own territory. It was responsible for almost half of Britain's trade. (source) It felt permanent. From the inside and the outside both.

It isn't there anymore.

Microsoft is not the East India Company, and I want to be clear about that too. But the pattern is recognizable: a company that grew so large, so dominant, so certain of its own position, that it started making decisions based on the assumption that its advantages were permanent. They aren't. No company's are.

The assumption that you can trade your people for infrastructure and the customers won't notice is the kind of decision a company makes when it's stopped being humble about what actually created its position in the first place.

Quick questions

Will AI replace my staff?

Not if you use it well. AI is good at repeat tasks like drafting, sorting, and summarizing. It does not build trust, notice a frustrated customer, or make the judgment call that keeps someone coming back. The goal is to free your people up for that work, not to remove them.

Does the personal touch still matter if customers can get fast answers from AI?

Yes. Speed helps, but customers remember how they were treated. When something goes wrong, they want a person who listens and fixes it. AI can handle the quick, simple questions so your team has more time for the moments that count.

Why does trust matter more than technology?

Anyone can buy the same software. Your competitors can, too. What they cannot copy is the relationship your team has built with your customers over time. That trust is what keeps people coming back, even when a cheaper or newer option shows up.

The lesson for everyone else

Most of us aren't Microsoft. We don't have billions to bet on anything. And honestly, that's an advantage.

Because the move that works at every scale, in every market, in every economic climate, is the one Microsoft is currently deprioritizing. Invest in people. Build a team that's clear on what they're doing and why. And use technology, including AI, to make their work better. Not to replace the need for them.

AI is a tool. A genuinely useful one. But it doesn't build trust with a customer. It doesn't notice when a client is frustrated before they say anything. It doesn't make the judgment call that keeps someone coming back for years. People do that. And the businesses that understand that, that use AI to free up their people for the things only people can do, are the ones building something that compounds over time.

Microsoft is betting that enough AI infrastructure will eventually force the market to come to them. The businesses that will win aren't waiting for that. They're investing in the people who are going to show up tomorrow and the day after, building relationships that no software upgrade can replicate.

Microsoft could still change course. They've done it before. There was a time when this company was genuinely people-first, and it showed in what they built. That version of Microsoft is worth rooting for. But the signal they're sending right now is worth learning from, whatever they decide.

Don't be Microsoft. Be the company your customers can't imagine leaving, because of how you made them feel, how you treated the people who served them, and how clear it was that you valued humans more than headlines.

If you want to think through how AI can support your team without replacing what makes your business worth coming back to, that's exactly what we do at Ascentis.

Sources: Microsoft, "FY26 Q4 earnings release"; TechCrunch, "Microsoft offers buyouts to around 60,000 employees"; Fortune, "Microsoft offers buyouts to nearly 30% of workforce in ongoing AI shift"; SiliconANGLE, "Microsoft offers voluntary buyouts to 60,000 employees amid AI investments"; Computerworld, "Microsoft to offer voluntary retirement buyouts to about 7% of the US workforce"; The Motley Fool, "Microsoft Q1 2027 earnings call transcript"; Data Center Dynamics, "Microsoft brought 88 data centers online in FY2026"; Directions on Microsoft, "Microsoft claims 15 million paid M365 Copilot seats"; UC Today, "Microsoft 365 Copilot passes 30 million paid seats"; Microsoft 365 for IT Pros, "FY26 Q4 Microsoft results"; SiliconANGLE, "Microsoft's stock rises on solid earnings and revenue beat, with Azure growing 48%"; Linuxiac, "Windows drops under 60% in global desktop OS share" (StatCounter data); National Geographic, "The British East India Company"