Untying the Gordian Knot of Cloud Spend with Wes Miller

Episode Summary

Wes Miller is a research analyst at Directions on Microsoft who’s focused on Microsoft identity, security, and management. Over the years, he’s worn many different hats, including serving as director of product management for Invodo and CoreTrace, a contributing editor for TechNet Magazine, an engineering manager at Pluck Corporation, and a product technology strategist at Winternals Software. He also did an eight-year stint at Microsoft, working as a product manager, program manager, implementation manager, and intranet web developer. Join Corey and Wes as they discuss the Directions on Microsoft origin story, the wild world of software licensing, how Azure currently plays second fiddle to AWS, how trying to figure out cost savings in the cloud is akin to untying a Gordian knot, Wes’ “cloud paradox” and what it means for predicting cloud spend, why Wes believes there should be a dedicated individual analyzing spend at a vendor when the numbers are high enough, why Microsoft is still the same old Microsoft despite what many people think, how marketing and developers often speak different languages, and more.

Episode Show Notes & Transcript

About Wes Miller

Wes Miller analyzes and writes about Azure infrastructure services, including Azure Virtual Machines and Azure Active Directory, and Microsoft systems management technologies.

Before joining Directions on Microsoft, Wes was a product manager and development manager for several Austin, TX, start-ups, including Winternals Software, acquired by 

Microsoft in 2006. Prior to that, Wes spent seven years at Microsoft working as a program manager in the Windows Core Operating System and MSN divisions.
Wes received a B.A. in psychology from the University of Alaska Fairbanks.



Announcer: Hello, and welcome to Screaming in the Cloud with your host, Cloud Economist Corey Quinn. This weekly show features conversations with people doing interesting work in the world of cloud, thoughtful commentary on the state of the technical world, and ridiculous titles for which Corey refuses to apologize. This is Screaming in the Cloud.

Corey: This episode is brought to you by DigitalOcean, the cloud provider that makes it easy for startups to deploy and scale modern web applications with, and this is important to me, no billing surprises. With simple, predictable pricing that’s flat across 12 global data center regions and UX developers around the world love, you can control your cloud infrastructure costs and have more time for your team to focus on growing your business. See what businesses are building on DigitalOcean and get started for free at do.co/screaming. That’s D-O-Dot-C-O-slash-screaming and my thanks to DigitalOcean for their continuing support of this ridiculous podcast.

Corey: This episode is sponsored by a personal favorite: Retool. Retool allows you to build fully functional tools for your business in hours, not days or weeks. No front end frameworks to figure out or access controls to manage; just ship the tools that will move your business forward fast. Okay, let's talk about what this really is. It's Visual Basic for interfaces. Say I needed a tool to, I don't know, assemble a whole bunch of links into a weekly sarcastic newsletter that I send to everyone. I can drag various components onto a canvas: buttons, checkboxes, tables, etc. Then I can wire all of those things up to queries with all kinds of different parameters, post, get, put, delete, etc. It all connects to virtually every database natively, or you can do what I did, and build a whole crap ton of lambda functions, shove them behind some API’s gateway and use that instead. It speaks MySQL, Postgres, Dynamo—not Route 53 in a notable oversight; but nothing's perfect. Any given component then lets me tell it which query to run when I invoke it. Then it lets me wire up all of those disparate APIs into sensible interfaces. And I don't know frontend; that's the most important part here: Retool is transformational for those of us who aren't front end types. It unlocks a capability I didn't have until I found this product. I honestly haven't been this enthusiastic about a tool for a long time. Sure they're sponsoring this, but I'm also a customer and a super happy one at that. Learn more and try it for free at retool.com/lastweekinaws. That's retool.com/lastweekinaws, and tell them Corey sent you because they are about to be hearing way more from me.

Corey: Welcome to Screaming in the Cloud. I'm Corey Quinn. I'm joined this week by Wes Miller, who is a research analyst at the interestingly named company, Directions on Microsoft. Wes, welcome to the show.

Wes: Thank you for having me, Corey.

Corey: Now, let's begin with the fact that I have absolutely no leg to stand on because I wound up once starting a newsletter called Last Week in AWS, and I've been dealing with the low-grade number of people who seem to think that I work for AWS, ever since. So, what is Directions on Microsoft, and how did you get there?

Wes: Sure. So, Directions on Microsoft is actually remarkably old. It’s, I think, 28 years old, And it was started by Rob Horwitz, in 1992. He started as a developer, and he went to business school, when he came back from business school, worked a little bit in marketing. And he just, sort of, came up with this realization that what developers were telling customers the software could do, and what marketing was telling the customers the software could do, neither one was really true. 

And so he wanted to split the middle. And he came up with this idea of building a newsletter, and he prototyped it internally. Threw it out and saw how many subscribers would be interested. Most of this was within Microsoft itself, and he got enough people interested that he basically self-bootstrapped it with a buddy that he'd met at business school, and we've been around ever since. 

And it actually started as Microsoft Directions and at the request of a certain local company, the name of the company changed, but it's actually been Directions on Microsoft almost since the beginning. And what we do is really explain to Microsoft's largest customers, partners, vendors, all sorts of people, what the company does, ideally where the roadmaps look like and help people in Microsoft’s sphere make decisions.

Corey: Excellent. I once had a Last Week in AWS affiliation on a badge at an AWS conference, and I show up; the people at the badging booth looked at Last Week at AWS presumably assumed that I worked there as the director of, “Take this job and shove it. My last day is Friday,” and gave me an employee lanyard, which was hilarious, but it's not the same level of confusion as from you, where people you talk about licensing if people believe that you are a Microsoft employee talking definitively about how licensing works, you'd almost be speaking ex cathedra, to some extent. It seems like the failure mode there is perilous.

Wes: Definitely. I think the reality is, a lot of people—first of all, a lot of people have never heard of us because—I wouldn't say we're boutique, but we're a small company. We're only interesting if you're a customer of a certain size in Microsoft's world.

Corey: What is that size, give or take?

Wes: Well, I can't pin down a number. But realistically, if you don't have, like, five people in your org, who are focused on Microsoft as a key part of their day job, it's not going to make sense. But you reach a certain point where if you're spending hundreds of thousands or millions of dollars per year, you'll realize, “Hey, it might be interesting to have an ally on my side who actually spends their day just focused on understanding the intricacies of Microsoft software,” And, as you and I've discussed, licensing.

Corey: So, one of my origin stories that I don't usually go into, but before I became a grumpy Unix systems administrator, which aged me 40 years overnight, I was a Windows admin for internal desktop support style stuff at a bunch of companies. And one of the things that drove me away from that early on was the joy of helping a company through a Microsoft licensing audit. And at the time, I had problems with authority because some things never change, and the problem that I experienced was I already had what felt to me at the time, like a very difficult job of making sure that all the computers kept working. But in addition to that, I had to sit down and become more or less an accountant and keep track of all the varying license arrangements and the rest. 

And I found this awful enough that it drove me into the world of Unix and Linux, and I, sort of, never went back after that. Looking back now, I marveled at how naive I was because we're talking basically a few Small Business Windows 2003 Servers, and maybe 20 or so desktop computers, and that's it. I didn't have problems back then. It felt like I had problems I didn't. At scale, this becomes an absolutely massive approach, but one thing that hasn't changed for me is that sense of why am I playing slap and tickle licensing agreement deals when there are actual, does the system do what I needed to do in a capability perspective? It always bothered me, and it felt like unnecessary busywork. Am I alone in that?

Wes: No, I think there's a certain point that a lot of businesses hit with if we go talk to a member of the press, or we talk to someone in the general public, and we explain what we do, and you say, “Well, I talk to customers who are spending—” and you show them your significant numbers of zeros, and they're just in awe. But these companies don't spend that money—I don’t want to say they don’t spend money carelessly, but they don't spend it trivially. They spend it because Windows Office and this infrastructure that Microsoft has created has become a key component of these businesses. And so the reality is that you keep spending that money in order to get the latest version of it to stay supported, to stay secure, and occasionally get some features, but in general, all that's being pulled towards services rather than software. But as a whole, I think you're absolutely right. 

The frustrating part of about this, both for me before I started here—I mean, I started working with SQL Server licensing in the last millennium, and it was not fun then and it's only gotten more complicated since then. For people who aren't in the business, the best way to visualize it is if you take accounting and chess and you combine them at high velocity, that's what software licensing is. It's what you get with Microsoft. It's what you get with Oracle. It's what you get with IBM. It's what you get with Corel, etcetera, etcetera. They all do the same thing. How can I make my software a revenue source that I can keep tweaking and modifying, tweaking and modifying and grow into new businesses, and keep my revenue going in a positive trend line? It’s what everybody does.

Corey: So, one of the interesting pieces that I'm seeing is whenever I talk to customers about their AWS bills, and oh, do I see enormous AWS bills, optimizing them becomes an exercise in planning, and prediction, in architecture. But licensing usually doesn't enter into it. The closest thing in AWS land to licensing concerns with multi-year concerns is things like commitments for spend, in varying ways: reserved instances, and savings plans just being two of them, but there are a lot of others. So, what I'm trying to wrap my head around those, whenever I have glanced into Azure bills, with customers who also have spent over in that side, the first time that happens, “Sure, I'd be thrilled to take a look and give you some thoughts.” And it was so intricately tied to custom enterprise agreements, there was license portability between what they had in their on-prem environment, and what they had in various cloud environments, the Software Assurance stuff added a whole nother level of complexity. 

And I quickly realized that A) I have no idea what I'm doing, and I can't responsibly respond to this, so I don't talk about it. Secondly, the problem that I was seeing was that there was so much complexity here that if I were to give advice on this without having either an attorney or someone who is very well versed in the intricacies of Microsoft licensing on my side, that there was going to be a great chance of me getting sued, and rightfully so. And thirdly, even if we do have someone like that on our side who gives absolutely correct advice, if Microsoft comes in and says, “No, your interpretation is wrong, you owe us more money instead.” Even if I'm completely right, we're going to be spending huge money litigating that if they decide to force the issue. Is my understanding of that correct? Am I basically sitting here fear-mongering without realizing it? I know that FUD has usually been the area that Microsoft liked to focus on circa 1996, but we're long past that. Am I spreading my own, now?

Wes: No, I think there's a lot of truth to it. So, for example, my colleague and I, Rob Horwitz, Rob spends a lot of his time focusing on programs like EA and other, basically, volume licensing programs. As I spent my day job focusing on technology around identity, and security, and systems management. When I talk about licensing, usually I'm talking about technologies like SQL Server, or services like Office 365 or, as you mentioned, Azure, which is this weird—I don't know the best way to describe it. It's just this cube, this complicated cube, which I'll circle back to. 

But I think what you've said is really important. You know, obviously, I'm pretty active on Twitter, and I answer questions related to licensing. Some of them are really gross, some of them are actually pretty trivial. The problem is that it's really easy, as we can see if you go look on Reddit, for licensing answers, any person who has a concept of what the answer could be will come in and chime in and say, “Oh, no, no, no, that's wrong. All you need to do is…” and beware of anybody who ever says, “All you need to do is…” because usually, that's wrong. The answer is always the most expensive. Whenever you're looking at whether it's Microsoft, IBM, Oracle, very rarely is the, “Oh, I think I can do this and save a little bit.” That's the thing that will come back and bite you. 

And I think you're exactly right with trying to analyze—this is where I think you and I have these different universes, and it's like comparing Marvel and DC in some way because you've got this whole sphere you're used to, and AWS I look at from the side, and I think I get these pieces, but there's a lot that I don't get. And then Microsoft, you're absolutely right that what the company has done—and we can look at a lot of the revenue growth and scratch that and say, “Well, that's very interesting,” But a lot of it is taking their on-premises advantage, and then using that to go towards the cloud. How can I get these Enterprise Agreement customers to spend a little bit on Azure, spend more on Azure? And a fair amount of this is back-scratching. In many ways, what Microsoft will do, it's not outright, but what you'll see is a little bit of discounting here in order for Azure spend there. So, it becomes a Gordian knot if you actually were to try and untie it because understanding what's Azure, what's volume licensing, and what's the spend inside of Azure, there's so many things moving at the same time, it's very hard to actually process and untie the savings.

Corey: And that's what gets strange for me from an analyst perspective—which is what I call myself when I want no one to know what I actually do, and cloud economist for whatever reason doesn't seem to fit the bill in that particular conversation—but we saw that Microsoft, at least as of the time of this recording, does not break out Azure revenues as a line item in it’s earnings reports. But they give percentage growth numbers, which is effectively useless in order to figure out how much it's actually making. Turns out it's super easy to get huge growth numbers on small numbers versus big numbers. So, first, it tells me that the numbers are probably something they don't want to disclose for a variety of reasons, but okay, great. 

It does make me wonder how much of Azure’s growth is effectively financial engineering of people signing agreements that have a portability benefit to them, expiring credits that are included that wind up being booked as revenue despite the fact that no workloads are being used that consume those credits. It really makes me wonder. Now, I do want to caveat this with I don't believe that the actual dollars going into any of the big cloud providers actually matters. If you're picking any one of the big three, you're not making a bad decision. I want to be very clear on that. So, who's further ahead and by how much? It's navel-gazing.

Wes: Absolutely. And I think especially when we're talking about number one, and number two, which no matter what anybody says to me, AWS is number one, and Azure is number two—

Corey: Fully agree.

Wes:—and it's because you have a first-mover advantage to begin with, but also AWS focused in on services first because that's where the company started. It was just one giant service. So, sort of stepping back and taking a look at Microsoft. One thing that I like to do when I'm looking at moves the company makes—and I do tend to apply chess metaphors to a lot of what the company does—everything operates on a triennium basis. It's a three-year basis because that's what an EA is in most cases. It can be longer in certain cases, but usually it's three years. 

And so, when you sign that EA, let's say, just for interesting scenarios, let's say June of 2020 because that's when a lot of them come up: at the end of Microsoft's fiscal year. You signed that, and it wraps back around until June of 2023. And so what are you agreeing to in that? You're agreeing to a bunch of things. You're agreeing usually to how widely you'll use Office, Windows, Client, System Center, maybe Microsoft Servers, Microsoft System Center, etcetera. 

So, you got all these things tied up in there, and the problem comes in when you start bringing in things like let's take two products, in particular, SQL Server and Windows Server because there's now benefits—they are literally called the Azure Hybrid Benefit for, and then insert product name. So, Azure Hybrid Benefit for Windows Server. And for Datacenter Edition of Windows server, you actually get a certain number of cores worth of Windows Server that you can run in Azure at no cost. It's not for free: it's actually a debit off of the cost you'd normally pay for pay as you go. So, it's a complex set of machinations. 

When you look at it, it's very interesting because it's an on-premises benefit of the software you license that's in Azure. And so the question I've always had is, “Okay, so if you have the rights to that, and you're always running it, and even if you didn't go in and check the box to say, use these cores, where's that accounting happening?” Because technically Windows Server sits in the Azure area. Your point is really valid that because you can't untangle it, and untwine it from Azure versus on-premises stuff, it's really hard to know what was sold, and what was spent, and who's using what versus, “Hey, we made a number go up.”

Corey: If you're like me, one of your favorite hobbies is screwing up CI/CD. Consider instead looking at CircleCI. Designed for modern software teams, CircleCI’s continuous integration and delivery platform helps developers push code with undeserved confidence. Companies of all shapes and sizes use CircleCI to take their software from bad idea to worse delivery, but do so quickly, safely, and at scale.

Visit circle.ci/screaming to learn why high-performing DevOps teams use CircleCI to automate and accelerate their CI/CD pipelines. Alternately, the best advertisement I can think of for CircleCI is to try to string together AWS’s CodeBuild/Deploy/Pipeline suite of services, but trust me, circle.ci/screaming is going to be a heck of a lot less painful and it's where you're ultimately going to end up anyway. 

Thanks again, to CircleCI for their support of this ridiculous podcast.

Corey: That is always the big challenge with any cloud provider, to be very clear. That, even without the license shenanigans, what's happening in the AWS space is at any non-trivially-sized enterprise where we're talking tens or hundreds of millions of dollars a year in spend, it's usually not one person spending all of that up, hopefully. So, it winds up being cross-division, cross-account, cross-functionally. And, as a result, when finance hears that, oh, the bill is 20 percent higher this month, is that the new normal? What does this mean for our projections? By the time that filters through corporate telephone, it translates into you're spending too much money. Stop it. 

And the big problem always comes down to it's not how much money, it's how do you attribute it? What caused this? Was it a fleet that was spun up that didn't get turned off? Was it a billing mistake on AWS aside—which never, ever happens, except when it does—or is it something else? What is it that drove that? The visibility into what is driving costs is super challenging. And strong credit were due to AWS, I have a crap ton of problems with how their billing system manifests in the real world, but everything that shows up on your bill, with remarkably few exceptions is you have a thing that is currently running. If you don't like paying for it, simply turn it off, and it goes away from that point forward. It doesn't feel like it works that way in a licensed software world.

Wes: It actually doesn't work that way in the actual licensed software world. And that's where Microsoft, sort of, gets the best of both worlds because we've got the company moving from a world where, again, everything's licensed on a three-year basis, and it doesn't matter whether it's shelfware, whether you buy it and you actually don't deploy it, which happens a lot with, like, Office 2019. We anticipate a lot of customers buying it, but sitting still on it because it's got such a limited support lifespan. And then, when you look at things like Azure, you're right. It's purely a services play and just—as I refer to it—Doctor Whovian pricing. It's really about space and time.

Corey: And the pricing is always bigger on the inside.

Wes: Exactly, exactly. And so if you're not using the thing, if you kill the VM—often Microsoft case if you spin it down and de-provision it correctly, you're not paying for it. You're obviously still paying for space, and space is often the thing we found will surprise customers. Like, “Wow, I totally didn't anticipate that A) that wouldn't go away and B) that we use as much space consumption on things as we actually wind up using.” To me, I think that's the fundamental problem with every cloud. 

And I think I've mentioned this to you before, that this thing that I refer to as the cloud paradox, the fact that if somebody comes to you and says, “How much will it cost me to do this thing in the cloud?” And you literally can't tell them. But what you can do is say, give it to me for a month, I'll run in the cloud. And I'll come back on the 31st, 32nd day and tell you how much it costs me to run it for a month. That isn't necessarily a good prediction of what it will cost to run for a year. I need it for a year to tell you that. 

And there's this whole problem, and so we wind up with—is exactly like you're saying, these projects that get spun up and spun out, and they're really broad across the organization because it’s not centralized, both in terms of development processes, and actually organization in our case, within the Azure world. People wind up spinning up things, either they forget to spin it down, or grows and become successful—Heaven forbid you actually used what you paid for—then you wind up with the opposite problem, which is how do I get cost control to work? And it's the same problem in Azure, AWS, I have to believe it's the same in Google Cloud, really anybody's cloud because you can't wrap it up and put it in a building, there's no way to actually go in and count it, and make sure that people are using things wisely, and not wasting money unless you really dedicate the resources to doing that. And we're just getting started as practitioners, or as teachers, trying to understand how to teach people to fish in this regard.

Corey: That is one of the biggest challenges. To some extent, there always becomes a question, at some point, of scale, where, when does this become an internalized core competency for any given company? So, one of my, I guess, strange questions for you is what is the value of what Directions on Microsoft does to company that’s spending giant piles of money on Microsoft every year? I mean, at some point, doesn't it make sense to have someone or several someone's whose full-time job is working through the intricacies of licensing?

Wes: Well, I'm obviously biased, but I absolutely say so. I think that the reality is, if you're spending a certain amount of money with any vendor, you should have somebody whose job is—a core competency is understanding the value that a vendor provides you, and really the negative value that a vendor provides you. Here's the tendencies of vendor X to try and extract more revenue from us year by year, so maybe we should be trying to control the spend on that. So, I guess when you look at what we do, there's a couple of things: in particular, focusing in on our update and our roadmap, we attempt to really take what Microsoft is doing, and distill it down so you can, as a reader who doesn't necessarily have time to digest all of these press releases and all of the things that the company is pushing out without press releases, understand really what's relevant to you, what the general direction is—we spend a lot of time on our roadmap to give you an idea of, every six months, here's what Microsoft 365 is going to look like; Azure is going to look like, and what but legacy on-premises software—and that's important: the legacy on-premises software because there's not a lot of love coming on-premises anymore—what are these things look like? 

And then the other piece is looking at licensing as a whole. We have our reference set, which is basically Wikipedia for people who have a licensing problem, as disturbing as this may sound. And the intent of all of it is to help you answer a question because we have these weird questions that people come up with, what's the answer? And there isn't a single great resource. And so that's what we've tried—we spent 12 plus years trying to create that, is to answer questions for people. And then, our boot camps are really about two things: teaching you how the rules work, so you can stand a chance of complying with them, and understand where they're going; and also, as you're, usually, an EA comes around, how can you maximize your investment that you're making, and make the best decisions for your business, which may or may not reflect the directions that Microsoft might want you to go.

Corey: And that's part of the challenge, is on some level, it feels like these cloud providers have all had the same problem, and I am sympathetic to this, where they want to push customers in a particular direction. And that doesn't always work because customers are where they are, and having a story of how they can get from where they are to a better place is all well and good, but it takes years, in some cases, for companies to even begin making moves in that direction. So, deprecation cycles become important. Microsoft has been terrific about having support for legacy things. And by legacy, I of course mean it makes money. 

Amazon, too, is famous for never turning things off. You can build businesses with assurance on top of virtually anything that they launch. Google is a separate story, but we'll get there down the road. The problem that I see with this is, when you're doing things like that, it's very difficult to get customers to move in a direction that aligns with your business when the strategy shifts. So, strategy shifts take, in some cases, decades at that scale.

Wes: Mm-hm. No, and that's what we see with legacy software is, again, when you look at what software people are running, I will often run into, on Twitter, people will say, “Well, why would a business buy Office 2019 and still be running 2013 or, Heaven forbid, Office 2010?” Because it's a sunk cost, and people don't get paid more money by rolling out new versions of software, necessarily. So, whether we're looking at Office or we're looking at moving from even in the cloud, from technology A to technology B or in Microsoft’s world, region A to region B where if you move to the new region, you can save money. It's crazy, but it's true. 

You have to strategically A) know that that's something you could do, B) know that it's something you, maybe, should do, and actually C) both take the initiative yourself, and get buy-in up the executive chain to say, “Why do we need to do this again?” And everybody along the way—no matter which org you look at today, in almost every organization, you're going to hit executives up there saying, “Why are we spending money on this again?” So, you, sort of, lose that will to push against it, and to try and make change. I think that's where we wind up with that stagnancy, that customers will sit still, whether it's on-premises, or they'll chuck it to the cloud, and leave it running in the cloud. 

And that's the thing that terrifies me about lift and shift, and Microsoft with their extended security updates for all the 2008 and 2008 R2 servers because you're reinforcing the worst habits. You're telling people, “You'll just take it from on-premises and put it in Azure, and we'll give you free ESUs.” And then people have maybe pulled it into Azure to do that, and they're getting free security for a maximum of three years, but what then? These are businesses that sat still for twelve years, or eight years. Do they have an actual plan to get off this software in the next three years? Because most of the cases, I would bet the answer's no.

Corey: One of the things that really rubbed me the wrong way—and I wonder how much of this is me just being hopelessly naive, is that last year, I was a bit of a champion for Microsoft. They invited me to the Microsoft Build conference—and again, turns out when you invite me, I wind up thinking more favorably towards you as a general rule, just because my standard rule of thumb is not I’m not sure how much I try and separate out editorial from other work and keeping it independent voice, it's very human that when someone does something nice and invites me, that I want to wind up treating them well. And it doesn't hurt that their products were legitimately awesome, and it was fun being able to look at what they're doing. I also have a personal standing rule of if you invite me to a thing that you're doing, and all I can do is trash you, I'm not going to go because that is unpleasant; it's not the brand I try to build. There are rules of snark, and that tends to run in the complete wrong direction. 

The challenge that I saw, though, is that they had this terrific story about their transformation. They can never quite come out and say it, but my strong belief toward the end of last year was that this is a new Microsoft. That the Microsoft that I hated, that drove me into using FreeBSD, and Linux, of all things, was dead and gone, and there's a new friendlier, cuddly Microsoft. And I believed that, and then they announced this whole Dedicated Instance licensing change at the end of last year, and I felt a little bit betrayed. First, can you explain what that license change was? And then let's talk about that.

Wes: Sure. And actually, I think it's really important to step all the way back and think about where this came from. The best example to understand is VDI: Virtual Desktop Infrastructure. If we look back years ago—first of all Microsoft's whole world, their whole realm is based on traditionally per-device licensing. And we've actually gotten to the point where cores come into play for servers, but it's really all about licensing the thing that that's going to run the software. And years ago, when Microsoft was selling Windows only on a per device basis, there really was no way to do VDI in the cloud. And Amazon came up with a couple of crafty ways with workspaces, one of which was to run Windows Server as a workstation, which is legally permissible. It's kind of funky, but if it does the job, who cares if you can make the revenue work, and it answers customer questions? 

But a lot of customers wanted to run Windows Client, and the deal was that if you wanted to run Windows Client on infrastructure, you had to have the hardware dedicated to you. And Amazon decided—they ran the numbers and figured out okay, well, what we'll do—and several other cloud vendors did the same thing—then what we'll do is we will lease the hardware to you, and it will only be your hardware, and because of the legal structure of that, you can now run Windows Client software on it as if it was your own hardware on-premises. And Microsoft has been pivoting and shifting a lot of the licensing rules over the past several years to honestly to monetize VDI, and we look at the whole Windows Virtual Desktop system they built now, you can see that direction. 

And so when we look back at last fall, there was this whole series of dedicated licensing changes. And what Microsoft did was, they laid down a series of rules that said, if you're doing dedicated hosting of Microsoft software, here's the new set of rules. And basically, they said, if you're a major cloud vendor, including Azure, it's no longer available to you. And the including Azure was, sort of, an asterisk because what they had done at the same time was add the Azure hybrid benefit, which didn't give the portability to other clouds, but did give Microsoft a bit of an advantage. So, I think it's important to look at your point that the new Microsoft—a lot of people talk about the new Microsoft. I'm famous for being cynical, as you may know, from following me on Twitter, and I think my challenge has always been that a lot of people are really bullish that the company has changed. 

The company hasn't really changed. What they've done is they've become much more strategic about messaging, much more strategic about open-source, and I think in general, a lot of this has been, for the customer, good. They've focused on open-source, they've opened up a lot of things. Not the key jewels, but they've opened up a lot of things. And that's an important piece because if you look at what they have open-source versus what they haven't, you can see the strategy inside of it. And it's the same thing with these dedicated rules, that what they did was they changed the rules in a way that, honestly it disadvantages every other cloud, and it advantages Azure. There's a distinct advantage to going to Azure now for Windows Virtual Desktop for your client, for Microsoft, for Windows Server, or SQL Server for your server operating system or your database. And it’s important because they did something that gives their own cloud a pretty significant advantage, and it's a forced move. 

A lot of customers can sit still in AWS, and AWS is working really hard to make sure that customers can continue to run the software as long as they believe it's permissible, and it is permissible for some period of time for some set of software licenses, but as you can see, with the report we published, the rules are very intricate and it affects everything from Windows Client, Windows Server, SQL Server, Office 365 Pro Plus—the Office client that you get as Software as a Service—as well as Office 2019 Professional Plus—the on-premises legacy perpetual software. So, there's all these things in motion. And every time we hold one of our boot camps every two months, we have to now explain to people usually it's, “Hey, we use this on AWS, what does this mean to us?” And I have to even stop and think, “Okay, you're using this variant of that, there. Here's your rules.” So, it added a lot of complications, especially if you're trying to use Microsoft software on AWS as a lot of customers still are, and will for the foreseeable future.

Corey: And that's the thing, too, is that very few people are migrating off of Microsoft for reasons that they used to. I don't get the stories of aggressive BSA audit enforcement the way that I did in the 90s and early 2000s. Is that gone, or is that just that they're getting better at shaping the narrative?

Wes: It's gone in the sense of—I think there are probably still BSA audits, it's not something I hear about. However, the way enterprise agreements work, agreeing to get audited is a part of the equation. So, it’s, by definition, it's there. The other thing that we are getting some sense of is that there's more interest in auditing the cloud, both directly and indirectly. And as a whole—like with some of the infrastructure Microsoft's put in place for SQL Server, and for the extended security updates in particular, in order to do that, you have to give Microsoft a fair amount of insight into what's running. So, in a way, you're giving them the ability to audit the un-auditable. Well, things that at least used to be something you couldn't audit, or audit easily.

Corey: And to be explicitly clear on this, people often lack nuance and context. I am in no circumstances advocating software piracy in any way, shape or form. That is not what this is about. But spending, effectively, person-months proving that you're in compliance when you've made a good faith effort to comply is what drives me nuts. I am not talking about lawbreaking; I am not talking about violating contract terms. I'm talking about people have very difficult jobs. We're all trying to do more with less, and spending all the time proving that your good faith efforts to comply got it right, is what's awful.

Wes: I totally agree, but one of the funny things is—so one of the sessions that I teach at our licensing boot camps, in addition to the Azure session, I teach our SQL Server session. And I always use an analogy from when I was growing up in Montana that we used to have a garden in the backyard. It was split between our house and the backyard neighbors. And every year we would try to cut back this one set of plots that were there so we could grow some strawberries and by, like, mid-summer, it was overgrown with rhubarb. No matter what you did, it was overgrown with rhubarb. 

And I actually always start with the same analogy for SQL Server because whether it's SQL, Oracle, or really anybody's database or anybody's office software, it's the same problem. It proliferates like rhubarb. So, if you don't actually have a process of constant consolidations—I was going to say remediation; in some ways it is—but constant consolidation for a lot of your key spend software, you're going to wind up with these problems. And so an audit will come along, and an audit will surprise you, and an audit will be very unpleasant. The other interesting thing that I feel is that a lot of bad licensing, again, it's one thing to understand what you're running, it's a whole other thing to say I'm actually living within the rules that Microsoft, or Corel, etcetera, provide or require. But it's a whole other thing to say we have no idea what software we're running because if you don't know the software you're running, you can't be license compliant.

Corey: Well, not with that attitude. No, I hear you.

Wes: Obviously, yeah. But the problem is, that also means that you can't be secure because if you have no idea what versions of software you're running, who's patching it? Who's maintaining your access control lists? It's a very similar problem, and in fact, I encourage people to have the security team and the asset management team working more closely together because they're actually working on a similar, or at least an interrelated problem.

Corey: That is, I guess, the challenge here. I mean, the last thing I'll say before we wind up wrapping up here, is that I have a lot of sympathy for many of my friends at Microsoft, specifically those who work on the Azure Advocates team. They work tirelessly to tell stories about how Azure can solve business problems, about the capabilities of the platform; they make the community better through their work, and I am in tremendous awe of what they're able to do every week. And when this change came out, one of my comments was that it, more or less, seemed like it turned much of what the team had said, into untruths or lies. And I don't blame the team at all. I guess I'm mostly just mad at myself for believing the transformation narrative without taking a more critical look to it. And I guess I'm curious as to whether or not you've seen anything in this space that would shed light on that.

Wes: You raise a key point, which is that a lot of those Azure advocates, a lot of the people I know as well, a lot of them are really good-hearted people who have the best intentions. And I think that, as a whole, everyone in Microsoft still wants to build software that makes the world more effective at what they do. And when we look at what Azure does, that's still the case. However, there's often this detachment from the people who are telling that marketing story—which Azure advocacy may not feel like marketing, but in some ways, it is because I'm helping you build a solution. That's actually where NT evangelization started, way back in the day. 

I think as we look forward, the reality is that people need to maintain that Microsoft is doing a couple of things at the same time, the first of which is, they're trying to make sure that Azure is the platform that people want to go to first. And part of that is by breaking norms that we would have expected in the past: you're just as Welcome to run Linux-based technology as you are Windows technology, but also that the company has levers that they need to push and pull. And at the end of the day, a key metric that the company both measures itself on and gets measured on by investors, and pundits is revenue. And how do you make revenue grow? It's by either inducing more customers or getting existing customers to pay more for either new technology or for the thing they already buy. So, the reality is, it's going to get more expensive to buy everything. It was E3; now it’s E5. For Azure, it was basic, now it's premium. These things will continue to change, and it's one of those things that I think you have to keep in mind that this person talking to you, they're an advocate, but they're not necessarily going to hold your business needs at the same tier that you need to.

Corey: That is probably one of the best ways to frame it. If people want to learn more about what you do, how you do it and hear the wise things that you say when opining on a wide variety of topics, where can they find you?

Wes: Well, they can learn more by going to directionsonmicrosoft.com. So, just directionsonmicrosoft.com

Corey: Yes. Motto, “We're not Microsoft.”

Wes: We're not Microsoft. Although you can call us for Directions to Microsoft, as people have done in the past. The staff loves that.

Corey: Indeed, it's the Microsoft equivalent to Apple Maps and Google Maps.

Wes: Absolutely. We’ll also tell you whether a Microsoft employee is a good hire or not.

Corey: No, that's Directors on Microsoft.

Wes: Oh, sorry, my bad. So, the other thing you can do if you're really feeling like it, is you can follow me on Twitter, I’m @getwired on Twitter, and I talk a little bit both about licensing, and politics, and a few other things. So, that's also one opportunity.

Corey: Excellent. Well, thank you so much for taking the time to speak with me today. I deeply appreciate it, and I feel like you've shed at least a little bit of light on a very dark and confusing place.

Wes: Thank you, Corey was great to be here.

Corey: Wes Miller, research analyst at Directions on Microsoft. I am Cloud Economist Corey Quinn, and this is Screaming in the Cloud. If you've enjoyed this podcast, please leave a five-star review on Apple Podcasts. If you've hated this podcast, please leave a five-star review on Apple Podcasts and then prepare for your licensing audit.

Announcer: This has been this week’s episode of Screaming in the Cloud. You can also find more Corey at ScreamingintheCloud.com, or wherever fine snark is sold.

This has been a HumblePod production. Stay humble.

Newsletter Footer

Get the Newsletter

Reach over 30,000 discerning engineers, managers, enthusiasts who actually care about the state of Amazon’s cloud ecosystems.

"*" indicates required fields

This field is for validation purposes and should be left unchanged.
Sponsor Icon Footer

Sponsor an Episode

Get your message in front of people who care enough to keep current about the cloud phenomenon and its business impacts.