
Loading summary
A
Foreign.
B
Whether you're an AI doubter or believer, it's undeniable that Microsoft is obsessed with AI. Employees know it, shareholders know it, Wall street knows it, and customers know it. But even customers who claim to see the wisdom of Microsoft's AI one mindedness are coming to realize that there is a finite pool of money, time and attention that Microsoft has to work with. That means something or things have to give. And if it's a Microsoft product or technology that your company depends on, that's getting short shrift. You may feel powerless, but you're not. There are some concrete things you can do. Even if you've decided, Microsoft may not care about what you and your org care about. We'll talk about the current situation and tactics on today's episode. Welcome to the Directions on Microsoft Briefing Podcast. I'm Mary Jo Foley, the editor in chief here at Directions. I'm your host for this series of podcasts for those interested in the Microsoft Enterprise IT ecosystem. Here today to talk more about his recently published blog post is Directions on Microsoft analyst Jim Gaynor. Jim leads the Directions on Microsoft editorial team and has been writing about technology since the early 1990s. Most recently, he's covered teams, Windows Server and related operating systems and enterprise IT infrastructure. Hi Jim, thanks so much for joining me today.
A
Always great to talk. I think the hard part when we get going is just staying on topic.
B
It is, that's true. Okay, so we talked a little ahead of today's recording and we said we don't want this to be a doom and gloom podcast. We don't want to be here to say AI is bad and you're bad if you think it's good. Not at all.
A
No. No. Yeah.
B
Right. So let's start out by talking realistically about what is happening inside Microsoft. And as 2026 begins, what from your perspective is going on with some of the teams, the products, and it's really not a surprise that Microsoft is all in on AI, Right? They've been claiming this for the past few years, but we want to know what's the cumulative effect of this AI or bus focus at Microsoft at this point.
A
So this is a podcast. I don't have a visual to show you, but there's a meme image that goes around. I'm sure you've seen it. And it's this guy with his head on a swivel turning to stare at a woman that's just past them while he's walking with his girlfriend and his girlfriend's looking at was shock and it's yeah, you've seen this, everybody's seen this one. And it's all about ignoring what you've got while you're distracted by someone else. And that's been happening at Microsoft with AI for years now, ever since they started started the big announcements at the beginning of 2023. You and I even wrote about a little back in 2024 when we talked about how teams, which Microsoft used to be all about promoting as the new platform for apps, was no longer the favored child. And that's happened with a lot of stuff. So a year and a half after that, Microsoft's this focus they've got on AI has led them to more or less ignoring a lot of the staid, boring, reliable things that companies run their businesses on. I mean you run your business on Windows Server, you run your business on intune, you run your business on all these boring management things that you can't always bolt copilot onto. And Microsoft until recently was trying to convince you to run your businesses on those. But now we've got this wave of retirements and consolidations of stuff I haven't seen in 10 years of covering Microsoft. We've got cancellation of once hyped previews, delays that are affecting widely used products like Windows Server Azure Local, which used to be Stack hci. The short run up that customers faced for moving to Exchange Server se slower updates to learn content. It all goes on. And then cumulatively it's not just that copilot and AI are the focus, it's that they're sucking the air out of the room for everything else. So after three years of that, it feels like a lot of the unsexy things that can't be hitched to AI that those bread and butter things I mentioned, they're gasping for air. They're turning blue.
B
Yep, agree. So in your recent blog post on this topic of Microsoft caring or not caring about customers priorities, you said this isn't the first time Microsoft has, quote bet the business on something they previously bet on the Internet and web and mobile. The latter bet not going quite so well. But what's different this time in your view?
A
Well, a couple of things. One is that what Microsoft is doing with AI requires a huge amount of capital, huge capital investment. You got to build data centers, got to populate them with high density services, expensive GPUs, costs a lot of money and that's a capital expense. That stuff that stays on the books has to be maintained. It depreciates over time. You're stuck with it. It's not like an operational expense, like hiring more developers. That's an operational expense. If you don't need them, you can lay them off. But once you've bought a data center, you've got a data center. So those capital expenses are a big thing. And Microsoft's been reducing all the operational expenses they can. All those layoffs they and other companies have been doing, reducing and because of the layoffs then doing that product consolidation and retirements I was mentioning earlier help offset all that money going into AI. This kind of feels, as I mentioned, it's big part of why these products are suffering from benign neglect or being pruned away because Microsoft is tremendously focused on AI. And there's also the fact that outside of certain areas there's some great things that AI, some things that AI is really great at. But in a lot of general areas it's been a push and not a pull. It's not making the massive general demand, the pull, the pull that we saw with web or with mobile or even older products like Windows 95. The general public, I'm not talking about businesses that are looking to optimize Six Sigma processes, but the general product wanted those things. They stood in line for those things. They were a pull. AI's got some great uses, but people are all. But your general public is seeing it alongside layoffs. They're seeing it bolted onto products they were already happy with. Executives are challenged to find firm, wide reaching ROI in their organization. And that's one of the questions we get asked a lot. By the way, how do I measure my ROI now that I've bought into this? And buying into it isn't cheap. But enterprise customers are kind of strapped right now. So again it's becoming more of a push from Microsoft rather than a pull. And that's also not a great thing. You've got that capital cost, you've got that pull, you know Satya saying, oh don't call it slop. And if people don't, you know, it's like that's not a good look.
B
No, no, no, it is not. And if you have to really go out of your way to convince somebody why they should buy something and you can't really tangible, easily demonstrable reasons, it's a hard sell as Microsoft's finding out. Right.
A
Well remember with mobile, I mean it was companies and I think we'll probably talk about this maybe a bit later but you had companies were trying to keep their people from using it. People were like they were sneaking things in the whole shadow it thing, which is still a thing, mind you, but it was shadow it, bringing things in from the ground up. And you're seeing that in some areas with AI, there's some things that it's great at and people are using it, but there's other areas where people are like, I don't want this, stop pushing it on me.
B
Yep, yep, indeed. So is it realistic to say, okay, customers you can just simply abstain, just stick with on prem and or cloud products that you know and need and ignore the AI centric wave of new deliverables, or do you really these days have to be a quote, frontier firm? Can you still be a Microsoft customer that isn't betting the firm on Microsoft's AI directions? Or if you do that, are you just a Persona on grata now?
A
So I want to mention really quick, the whole frontier firm thing, I just.
B
Yeah, it's a weird.
A
Well, they treat that like it's like, oh, it's some industry term. Like Microsoft made up the term frontier firm.
B
Yeah.
A
The first time, the first reference was a study was they did a white paper like a year and a half ago and that was the first time they mentioned frontier firms. And now they're mentioning it as if it's something. And I just find that amusing. Although they talk about frontier models in AI and that's probably where you got it. But the FOMO is real. The FOMO is real. Pressure's everywhere. I mean, even if you're a C suite executive who's made your decision that you haven't bought in, you've probably got pressure from your board members and from your investors, let alone if you're a mid level person in the offices. And so at the moment there's a certain amount of it being inescapable, but you can minimize the exposure in the same way that you could with effort, hold off on the move to the cloud when that was inescapable. And most people have moved to it somewhat. But there was a time when it's not right for me now it's not right for me in this shape. And it takes that same kind of diligence. You've got to be focused on what you want for your company, what you think works, and then manage things. Otherwise you need to look at Microsoft's announcements with a critical eye and realize just how much of what they announce is aspirational and early previews to sift out what's actually a shipping product with the impact and support and controls you need versus something that they're trying to get you to buy a license for now, but they won't actually give you a product for, you know, maybe next year. And like I was talking about shadow it earlier with mobile devices or cloud services like Dropbox Renew, there was shadow it you had to worry about to keep people from using them and creating issues. And you do have to worry about that now with your employees because for some areas it's really good, helpful and consumers, people are using consumer AI in the workplace to do what they need. You've got to worry about that there will be, will be employees trying to sneak it in, but also Microsoft's slipping things into your existing products to try to create that pull. I mentioned earlier. It's a classic tactic. I remember them doing it with Exchange back in the, back in the early 2000s. So you've got to watch out for that and give any kind of controls that you need to prevent those things from popping in and saying, well, if you just take this upgrade, you'll get this and then you've got this clamor going on. So you don't have to be a frontier firm, but there's effort and diligent and diligence involved because Microsoft wants to make that pull and to do it, they're creating, they're trying to create a current around their products and you've got to kind of swim against that.
B
Yep. I'll tell you just as a personal anecdote, Notepad, how and why they're adding AI to Notepad, I do not understand it, but luckily you can turn it off again, you have to be diligent, which I am. But yeah, yeah.
A
Oh, when I saw that, when I first saw it, I'm like, oh, Mary Jo is going to be pissed.
B
All right. On that note, talking about my favorite Microsoft product, let's take a quick break so I can talk about why you should make it a priority this year to attend one of our directions on Microsoft licensing and negotiation boot camps. As of November 1st last year, Microsoft eliminated tiered volume discount for online services in the Enterprise Agreement and mca. Unless you were and are prepared, this change means you are going to be facing a potential 6 to 12% price increase on your core Microsoft Cloud services like Microsoft 365, Azure and Dynamics 365. So what can you do? Your best defense is to invest in your Microsoft licensing expertise. Our next in person directions on Microsoft boot camp is happening very soon in Los Angeles, February 3rd to 5th. You should send your IT procurement and ITM leaders to learn how to deal with Microsoft's new Licensing reality. You will learn things there like how to control your Azure consumption costs so you can cut your largest Microsoft cloud spend component. We'll help you do things like master the strict licensing rules for high cost products like Copilot and Microsoft 365 add ons so you can ensure compliance and maximize your organization's ROI without overbuying. And we'll even help you craft a tailored framework framework for your renewal strategy to secure concessions that offset lost volume discounts. If you want to secure your spot today, go to directions on Microsoft.com training. You'll get to talk to people like Wes Miller, Rob Horowitz and our other Directions licensing experts there. And if you can't make it to Feb to LA in February, we do have other boot camps coming up later this year in person in Washington D.C. and Chicago too. Okay, back to Jim and Microsoft's AI obsessions. Now let's talk about tactics. We want to offer Microsoft customers who feel like Microsoft no longer cares about what they care about some real things they can do. And there are some steps you can take to try to get what your business wants and needs right. Just idly threatening to quit and going to Google Cloud or AWS isn't the way, given those companies are pretty heavily focused on AI too. Not to mention technical complexities of quote quitting Microsoft. But what are you telling people they could try?
A
Jim so it's honestly this isn't a cross sell, but since you mentioned the boot camps, I mean a lot of the tactics for this stay the same whenever you want what Microsoft isn't selling at the moment, the best things I can do are summarize some of the points that Dean Bedwell, who leans our advice, who leads our advisory group often makes. And I've just had to internalize them because yeah, they work, they're good. When you're trying to get concessions for from Microsoft on contracts either to say hey I don't want this or I do want this. It always helps not to be all in on Microsoft. You mentioned maybe you're not gonna go to Google Cloud, you're not gonna go to aws, but keep your fingers in those competitor pies. You may not wanna go there, but if you use some in your company, you've got some Amazon, you've got some Google. Then the threat of your salesperson of you know we're thinking about expanding our relationship with Amazon or Google, well they're gonna be a little bit more anxious. You know, they wanted to keep, you know, we want your new sales to come from us, not them. So that's really an important thing to do, have those options there. Also, Microsoft salespeople are given targets that are compensation depends on. And right now that's all AI. I mean, you saw it. You snuck into some things at recent shows and you saw that. So giving them a little because you know, they're salespeople, they want to hit their metrics so they can get their bonuses can go a long way for you. You know, give a little, get a little, maybe buy some more Microsoft 365 copilot than you planned, maybe, you know, buy into Azure AI Foundry and in exchange for concessions on things you really do need, like Maybe you want ESUs on that version of Windows Server that's still keeping this legacy database up and going. Maybe you want special use cases that require licensing exceptions or credits for services that your business already relies on. Extended support, not the ESU kind, but more support from Microsoft on helping you migrate. All of those things you can hopefully do that will help you if you give them a little bit more. The thing is though, is you can't really get Microsoft to suddenly pay attention to the product or service you want them to, but you can at least find some leverage to make it easier to stay on what works for you. And I know I've harped on this already, but that's really the important thing is you've got to stay focused on what your business needs to thrive and what Microsoft actually has in hand that works for you rather than previews and promises.
B
Okay, those are all great points. Excellent. Okay, as you mentioned earlier, FOMO fear of missing out is real when it comes to AI. But given the pace at which things are changing, especially when it comes to Microsoft and its AI services and products, being first seems riskier than waiting for the dust to settle at this point. What do you think?
A
Yeah. So like we've said, there's scenarios where what's available now works and can have a lot of benefit. I mean, we're really seeing a lot of good stuff for AI being used in development, for prototyping, for running drafts, for getting summaries of content. And these are places where it's generally useful now. And that's what you got to focus on. Focus on what works for you and works for you now, not out of fomo. Because if you're adopting because you. Because I need to be first. Well, you probably don't, and that can hurt you. One example I'll give is Model Context Protocol mcp. So this is a protocol that gives A standard way for AI to retrieve information from another source. It's what's called Retrieval Augmented Generation or rag. And that was a big deal. Like two years ago, everybody was like, oh, retrieval augmented generation. Because that will allow me to have my AI query, my support, my library of support content to make the responses better. But everybody's having to custom develop that. Well, then MCP comes along. It only really became a thing about a year ago when Anthropic released it to released it publicly. And it streamlined what had been a laborious bespoke process to set up to get that generation to import that stuff into your LLM. So if you'd invested heavily in custom developed retrieval augmented RAG pipelines before then MCP kind of made that time and investment effort pointless. So I'm not saying that you shouldn't have done that. If it was something that your business depended upon doing quickly, then by all means. But if you're just doing it like, oh, we've got to have one too, you might have just wasted a bunch of time and money when something simpler was being developed. Here at Directions, we're analysts, so we do some news, but breaking news isn't our focus. And yet every once in a while, we kind of all get all frothed up about something. Oh, this is news. We got to talk about it immediately and kind of, you know. And you've been in those meetings.
B
Guilty as charged right here.
A
Oh, yeah, we all do it. We all, we get excited, but then it's kind of, you know, to calm back. I kind of remind myself that we don't want to be first, we want to be right.
B
Yep.
A
We aren't. Our focus isn't news. Our focus is making good information available. Again, talking about focus, not. Not puffing us up. So we focus on taking the time to be right. And I think that same kind of focus applies here too. What do you need to do? What's important to you? Are you able to take the time to do it right? And so sometimes waiting a little bit for tools that can make things more appropriate for you, more useful for you, let things shake out. Like some of the, like Microsoft keeps changing things. Letting that shake out can really help out.
B
Yeah, for sure. All right, I'm going to throw a surprise question in here because I know this is a topic that's probably near and dear to your heart. Just guessing. So recently Apple and Google announced an AI alliance. So I'm curious if you think this poses a risk for Microsoft and its customers in any way, or is this just Apple playing AI catch up.
A
It's not any one of those, I'd say on that one. I mean, I don't think it poses a risk to Microsoft and its customers. It is Apple paying catch up because I mean, countless analyst words have been spent on how Apple bobbled Apple Intelligence almost two years ago now. But what's interesting about this one is Google has mostly stayed out of the whole Microsoft, Oracle, OpenAI, Nvidia thing. I mean, they had their own AI going, they kind of held back and OpenAI stole the thunder when ChatGPT 3.5 came out and now they're catching up. But Google, interestingly, they're developing their own models, they're developing their own stuff and all of their stuff, all their inference, all their training is being done on their own silicon. They've got their own TPUs, the tensor processing units. They're not using, you know, Nvidia, Blackwell GPUs. And then you've got Apple who similarly, when they talked about Apple Intelligence and what they have tried to do so far has also been focusing on their custom Apple silicon and the neural processing units that they've developed inside of those. So this is interesting because it, it, it's a play for AI A, on a more consumer basis because that's what Apple works on. They don't really care much about enterprise, but it's outside of that circle of companies that people have talked about before. Microsoft, Oracle, OpenAI, Nvidia. And so it'll be interesting to see where that goes. It's also interesting, it's also a good thing, frankly, in my opinion, for Microsoft because you've seen this Microsoft, when, once they, once Microsoft owns a market you don't want, it's like if things get weird, things get unfocused, things get slow. When they've got a. Yeah. When they've got a competitor, when they've got somebody that they think is a threat of any, they love to think of themselves as the scrappy underdog. And so if they've got someone to compete with their own stuff gets better.
B
That's true, very true. Yep.
A
So if you, if we can have some players, some meaningful players, opposed or not even opposed, but just not working in the same way that Microsoft is working, that Microsoft can allow themselves to see as a competitor, it'll be better for everybody involved. Well, maybe not for them involved, but for all of us who are trying to use this stuff. They'll be making better products.
B
Right, right. I always think more competition is better, not just for customers, but like you said for Microsoft itself.
A
All right, you don't want a monoculture. You don't want everything all the same.
B
No, you do not. All right. Well, on that note, Jim, I'm going to say thank you so much and that was a really fun podcast. Thanks for joining.
A
No, like I said, always a pleasure. Love it. Good.
B
I would like to remind our listeners they can find lots more coverage of all things Microsoft related on directions on Microsoft.com thanks again for listening. If you have questions, comments or any topics you would like to hear the Directions analysts cover in one of these podcasts, please do not hesitate to contact me via X or BlueSky. Directions on Microsoft also is on LinkedIn. Make sure you follow us there and give us follow at DirectionsMSFT on X or directions on Microsoft on BlueSky for all of the latest Microsoft Enterprise product and licensing information. Thanks again,
A
Sam.
Podcast: The Directions on Microsoft Briefing Podcast
Episode: "Microsoft Ignoring Your Org's Needs? We Have Some Tips for That"
Date: January 23, 2026
Host: Mary Jo Foley
Guest: Jim Gaynor, Analyst at Directions on Microsoft
In this episode, Mary Jo Foley and Jim Gaynor discuss Microsoft’s recent organizational and strategic shift towards AI, exploring how it’s impacting longstanding products and what enterprise customers can do when they feel left behind. The tone is pragmatic and insightful, blending industry context, customer strategies, and real-world examples for navigating Microsoft’s changing priorities.
Microsoft’s overwhelming focus on AI is leaving some enterprise customers feeling neglected, especially those relying on "unsexy" but critical on-prem and core cloud products. The episode explores the consequences of this shift and practical tactics for customers seeking attention for their own priorities, despite feeling sidelined by Microsoft’s roadmap.