Save to My DOJO
Long ago, near the beginning of my technical career, I worked for a company that serviced a niche PC peripheral. One customer’s system behaved so poorly that she needed a service call almost weekly. Once, a frustrated technician asked, “Our competition sells better products, why do you stay with us?” Her response: “You show up.” Few lessons in my career have served me as well as that one. With that frame of reference, I will explain where I believe that Microsoft fails its customers and ultimately itself.
The Face of Microsoft
Hopefully, most of you have watched videos from Ignite and similar resources since they became freely available online. Obviously, the high energy and dazzling polish of the presenters and their material originated with intentional design and practised delivery.
However, since most of you never interact with the product teams that create and build Microsoft’s products, let me tell you that most of them bring that same energy to their work. I have seen PMs light up like children with new toys while talking about innovations and advancements in their products.
Whatever you might think about the Microsoft Corporation, they actively and deliberately push technology forward. Some of the world’s most passionate, intelligent, and talented people wear Microsoft’s signature blue badge. They genuinely want to solve the technical problems of today and tomorrow. Even though I believe that the company misses many vital marks, it would be inappropriate to blame much of it on a lack of ability or interest. Their problems come from more innocent sources.
Big Buyer Blinders
I once overheard a couple of salespeople discussing their accounts. The junior made a comment about focusing on his bigger clients due to the balance between the level of effort and reward. He posited that if he spent the same amount of time on a large client versus a small one, he would get more from the larger business. The senior salesperson responded, “The big customers put the Ferrari in your driveway, but the little customers put the meat on your table.” He went on to explain how smaller customers tended to have greater loyalty, the people you interact with at small companies often become the people that call you when they move to big companies, and a few small clients switching to the competition don’t have the same catastrophic effects on career and commissions as the departure of one or two large customers.
We can extract several morals from that exchange. Let’s tease out one: small customers matter.
The Boundaries of Technology
We all know stories about products with superior technology that lost to an inferior offering. The classic Betamax vs VHS battle leads the pack. In my experience, the vendor that holds the greatest market share of line-of-business software for any given vertical typically inflicts an excruciating experience on its customers. They have multiple competitors with no other goal than doing a better job. Somehow, those products never manage to shift the balance regardless of technical superiority.
Hordes of people with better analytical skills than I have investigated these cases and drawn all sorts of conclusions. I want to focus on what my customer from the opening paragraph was telling us and the lesson that I took from it. While she could have easily purchased a more reliable system, she knew that it would still break. When that happened, other vendors might take weeks or even months to respond, and she found that they sent sullen, impolite technicians. Conversely, we arrived onsite within a couple of hours and treated her pleasantly. Our “inferior” product provided substantially greater net uptime and the customer enjoyed interacting with us.
From all this, we learn that, while technological innovation and features matter, it takes more to satisfy customers’ needs.
Microsoft’s Major Misses
I believe that Microsoft has not absorbed either of the lessons from the previous two sections. Microsoft will survive, of course, but it constricts its ability to grow and connect with customers.
The Customer is… Something
Think about the large technology names that have held dominant positions for decades. When you consider your own experiences, how do you judge your interactions with them? Set aside the products for this mental exercise and only ponder your customer experience. Do you feel like a name or an account number? Outside of any personal connections, how do you feel about the individuals whose names you know in the organization? Would you characterize their attitude toward you as more on the caring side or more on the condescending side? If you have worked for employers of different sizes, do your answers change in correlation with the size of the firm?
While you ponder that, I’ll summarize what I have seen. The old guard technology corporations with nothing to fear display almost no concern for their customers. They might occasionally publish prepared executive statements and infrequent marketing platitudes, but little else. The smaller your business, the less love you receive. If you have only spoken with their support departments, you likely have a cruder opinion. With tongue firmly in cheek, I proclaim that a system administrator graduates from “junior” to “senior” the first time they endure abusive treatment from support on a call where the technician knows less than the administrator.
Between my experiences and those gathered from other MVPs and server administrators, Microsoft behaves mostly like the rest of its cohort. I have had some wonderful interactions with Microsoft support, but the balance tallies to an overwhelmingly net negative. Furthermore, all my positive experiences occurred while working for organizations with deep pockets. Like so many of you that reached out to me from smaller companies, I too have felt that I was setting my support dollars on fire and getting about as much value from my time as a losing game of solitaire.
Look at this Shiny! Wait Look at that Shiny! Hold On, It’s Another Shiny!
Fortunately, almost all of us eventually find solutions to problems with Microsoft software even if we only find answers when we look outside of Microsoft. They make a larger problem in the way that they prioritize technological development.
You don’t need any special connection to Microsoft to see their direction. They keep silent on individual features until the last minute like any prudent software company, but they paint their roadmaps on giant canvasses with large brush strokes. Due to Microsoft’s size, what comes to mind for you has a lot to do with your focal area. For me, I see many things about Kubernetes on Azure, Kubernetes on Azure Stack HCI, tools for Kubernetes, deployment for Kubernetes, etc. I don’t need to have a single interaction with Microsoft employees to figure out that they have poured resources into container infrastructure.
At first glance, that doesn’t look like a problem. Microsoft’s core business is software development, they operate one of the world’s largest public clouds, and they provide the most popular general-purpose server operating system. If you follow or participate in any development communities, you can’t miss the enthusiasm for the promise of container technology. If you watch the direction of technology in large businesses, you see how many of them have adopted containers or have them in their plan. From that viewpoint, I understand Microsoft’s interest.
Now, recall the earlier discussion about customers. Shift your view frame from developers and hot trends to daily business, especially the smaller ones. I do not have a single container in production. No one that I talked to for this article had a container in production. I could not find any evidence of container production by any of the dominant line-of-business software vendors in any vertical that intersects my work. I can find plenty of evidence of container use, but not in a way that drives daily business. Every time I meet a container advocate, I challenge them to make containers relevant to what I do today. They always fail.
As previously stated, it makes sense for Microsoft to get ahead of the game. They do right to take containers seriously. They go wrong in setting aside so many urgent things in their race to the future. They do that frequently.
Need other examples? Microsoft’s server teams have worked tirelessly to make in-place upgrades flawless. However, only an insignificant portion of administrators ever perform in-place upgrades because even if the risk declines to effectively zero, the payoff still does not make it worth the effort. What about Windows Admin Center? Microsoft has kept that front-and-center for Windows Server administrators and adds features constantly. The problem? They generally skip over the larger pain points of administrators to introduce flashy new gadgetry. In probably the ultimate example, I performed an in-place upgrade of a server running Windows Admin Center from Windows Server 2016 to 2019 and it wiped out every administrator’s WAC configuration. Two of Microsoft’s favorite toys burned in one fire.
Lest I let it appear that I have a few pet peeves and a handful of personal experiences clouding my vision, let’s look at some other places that Microsoft had its developers fiddle while Rome burned:
- Guarded Fabric and Shielded Virtual Machines (although, judging by the date and content of a promotional article, not even everyone at Microsoft got that memo): In almost all cases, the organization trusts its hypervisor administrators with the safety and care of virtual machines. What existent problem did Shielded VMs solve? I recall several hosting providers diving into it, but most of the rest was more casual interest. As for those hosting providers, they seem much less enthusiastic about Microsoft in 2022.
- Semi-Annual Channel releases: Few administrators have taken to the Core mode of Windows Server and almost zero line-of-business applications support it. The thought of an 18-month support cycle on a server operating system makes CIOs and IT directors panic. What problem did SAC solve?
- Nano Server: Nano Server sounded cool but ultimately, it couldn’t solve whatever problem it was designed to address. If you missed that hype, the net effect was that starting from Nano and making it useful just left you with something that looked a lot like Windows Server in a Core installation.
- Hyper-V Server (the free standalone SKU, not the hypervisor technology): Hyper-V Server was awesome and amazing and solved lots of problems for administrators and introduced new opportunities, but Microsoft wanted personnel working on containers.
- Office edition of OneNote: The desktop edition of OneNote was a wonderful collaboration tool, especially for administrators storing notes for themselves or others as a rainy-day solution, but Microsoft wanted an Azure-only solution that does not support open-ended sharing and is unreachable during the sort of outages that send administrators to their notes (but hey, it’s free now).
- Windows Phone: Windows Phone had adoption problems and Microsoft should never have purchased Nokia, but the platform deserved a chance. An easy-to-use phone that literally anyone could develop for using easily learned free tools? Windows Phone was everything that Android and iOS proclaim to be. It only lacked a big app store and sufficiently passionate leadership.
- Visual Basic [.Net]: Visual Basic in either incarnation was a fantastic language and environment. Non-programmers could rapidly put together all sorts of nifty tools for themselves and users. Kids could learn it quickly. It could create software for the desktop, the server, and mobile. It’s the only language I’ve ever used where I could return to a project after six months and be able to immediately pick up where I left off. But, some loudmouthed people on the Internet don’t like VB because of inconsequential reasons so Microsoft effectively killed it for everyone.
- Dark mode: I know how popular dark mode has become. Not only do people love it, but they also love to gatekeep and shame anyone that won’t jump on the dark mode bandwagon. The thing is, 100% of scientific studies comparing light vs. dark mode concluded that light mode is superior in every metric except subjective preference. This obsession has not been a good use of anyone’s time, particularly a company responsible for some of computing’s most widely used interfaces.
- Honorable mention to Hyper-V (the hypervisor): Hyper-V is an outstanding hypervisor and it powers Azure, so it doesn’t exactly belong on the list. However, a lot of people don’t realize how well Hyper-V can stand up against other hypervisors because the dominant tool to manage it, System Center Virtual Machine Manager, is… well, it’s not great. It’s “not great” enough to actively turn people away from Hyper-V.
With a little brainstorming, we could come up with an extensive list of products and technologies that Microsoft bet on and lost. We could do the same for any company, though. The gambling part does not concern me. It’s the pattern that I see when I review that list: Microsoft expends exorbitant resources on solutions that don’t have problems while tossing out some of their best work. Their behavior reminds me of a cat enthusiastically but fruitlessly chasing a laser pointer.
The Squeaky Wheel Gets the Grease, the Broken Wheel Gets Nothing
Microsoft doesn’t only get distracted by the playground. Although many of you never experience it, they do maintain close contact with many customers and take their complaints seriously. So far, every Microsoft employee that I’ve ever heard start a sentence with, “All the customers I talk to say…” go on to finish that sentence with something that doesn’t impact small customers. As an MVP, the bulk of my Microsoft contacts hold important leadership positions. So, that means that they can push for changes and fixes, but the problems that reach them as a priority do not come from sources that fairly represent Microsoft’s customer base.
Additionally, how many enduring problems irritated you for months or years in Microsoft products? How often do you receive error messages that give no indication of what to do next? I could list several, but I want to call out one that I encountered in Azure so that I can refer to it later: When using my crummy, slow Internet connection with a large file that reaches the timeout period, Azure File Copy and all its various incarnations refuse to honor the resume features. When it fails, I get an error “message” that tells me nothing at all. My Azure account gets charged for the space of the entire file, though. To tie this problem back to the support subject, I found one person at Microsoft that really wanted to help me but did not have sufficient product ownership. Everyone else that I tried or that he referred me to did not consider me or my problem important enough to even acknowledge. So, I can’t upload a big file to Azure, but if I want to set up a container that won’t do anything useful, then I have more resources than I know what to do with. As a quick mental exercise, imagine the ratio of the number of people that want to move files against the number of people that want to run containers.
The Coming Consequences
From what I’ve seen from Microsoft, I doubt that they deliberately created these conditions. Microsoft grew organically and has experienced major mutations, especially across the lengthy transition from the Gates era to the present reign of Nadella. A quick review of their stock price history shows that they have grown consistently, so maybe they feel like they can afford to continue as they always have. While I don’t believe Microsoft will put itself out of business any time soon regardless of what they do, I believe it could do far better by getting out of its own way.
I feel that Microsoft’s recent successes have obscured many problems. Everyone has realized that Microsoft has dramatically slowed their on-premises products in favor of Azure. Given what the cloud has done for their bottom line, no one can fault them. However, consider this:
- Much Azure adoption came organically from companies with heavy utilization of Microsoft products on-premises
- Microsoft does not lead the cloud computing industry
Currently, Amazon is the biggest name in cloud computing. Microsoft seems to consider them as just another competitor to push aside, like Novell. However, Amazon brings something new. The big names that Microsoft effectively retired in the past behaved like everyone else in that cohort. Conversely, if you live in any part of the world where Amazon has a major presence, then you or someone that you know has at least one story of Amazon customer service going above and beyond. Tally them up against similar stories about Microsoft or any other multi-decade fixture in technology and let me know how that ratio turns up.
Let’s summarize the implications:
- A substantial number of administrators have experienced Microsoft’s failure to support their on-premises datacenters, so they have concrete reasons to prevent Microsoft from becoming their datacenter
- Amazon has positive brand association and customer loyalty built up among people that will start new businesses and promote into leadership roles in existing businesses
- Allowing the small things to remain broken encourages people to move the small things elsewhere, which generates incentives to continue moving things
Technology does not appear on that list because it does not matter. Azure has astounding features and that doesn’t count the things that my NDA prevents me from telling you about. However, most people don’t need any of it. They need technology that solves the problems that they have. Look at my situation: I would like to use several different Azure technologies, but I can’t upload a file. My journey stops there. All the things that depended on that file upload will go somewhere else (truthfully, in my case, I just had some cool ideas for an article that will never see the light of day). If I were a CIO or small business owner, then I would probably start thinking of the non-Azure solution first when I start or refresh a cloud project because I already had to take something else there. No one ever does a complete cloud lift-and-shift, but gradual migrations start like this.
Righting the Ship
Whether I’m right or wrong on my interpretation of current events, it still stands that Microsoft devotes its talents to building hammers that spend their short lives desperately searching for nails while restricting the resources at work at plugging the leaks springing up everywhere. Unlike in the past, Microsoft has competitors that take a different approach that appeals to modern businesses in a way that Microsoft does not. Fortunately, they have plenty of time to make it right.
A few suggestions to Microsoft:
- Listen to your small customers. Seek them out. Re-open UserVoice or institute a viable substitute. The Feedback Hub is not doing what it should.
- Fix your support department and processes. It would help to make it more integral to the business and less of a walled garden.
- Task resources on every single project to create useful and helpful error messages. “Something went wrong” was not improved when it became, “Sorry, something went wrong”. I would bet that forcing developers to implement valuable error messaging will help find a lot of these bugs that keep sneaking into production software.
- Do not direct all your best talent to the latest fashion. Your customers need you to maintain strong focus on the problems of today. It does you no good to be at the forefront of tomorrow if all your customers left to find other solutions.
I will let my customer give the final, best piece of advice: “Show up”.
Not a DOJO Member yet?
Join thousands of other IT pros and receive a weekly roundup email with the latest content & updates!