X
Business

What Microsoft does well

I asked the question yesterday: what does Microsoft do well? The question is predicated on the assumption that every company with a reasonable degree of success has something it does that is unique, and that was instrumental in its success.
Written by John Carroll, Contributor

I asked the question yesterday: what does Microsoft do well? The question is predicated on the assumption that every company with a reasonable degree of success has something it does that is unique, and that was instrumental in its success. In Apple's case, it is their design skills in both hardware and user interfaces. Good design would rightly be considered part of Apple's core identity. A utilitarian gray plastic device with all the aesthetics of a screwdriver is as likely in the Apple line-up as a lawnmower section at a BMW dealership. The core identity, in other words, helps to stamp out deviation from what is, essentially, a very good aspect of Apple products.

Core identities, when properly understood and allowed to permeate a corporate culture, act like a constitution. Constitutions tend to lay the broad outlines of what is considered proper within the legal and political realm. It serves as a strong curb on those who wish to deviate from base principles, and more specifically, trims what people with an interest in power are able to do in pursuit of their interests.

Questions of identity are particularly relevant to Microsoft at this juncture. Microsoft is contemplating a merger with a company that has 12,000 employees, not to mention a Unix development culture that is, for obvious reasons, quite different from that which exists at Microsoft. Microsoft is also of a size that it is harder to discern a unifying theme across its product lines. Microsoft makes operating systems, set-top boxes, databases, Customer Relationship Management (CRM) software, music players, game consoles, and interactive touch-sensitive tables, to name just a few from a list of products that would simply be too long to mention all at once in a blog post.

It is also a company with famous levels of intra-company politics. Microsoft isn't unique in that. Big companies with lots of money are attractive to people with a skill for and interest in power. No company in the software industry, however, is as rich and powerful as Microsoft.

In most IT organizations, code is power. That was the case at a bank I briefly worked at prior to Microsoft. Some team created a convoluted time entry system that they managed, through political connections, to mandate that everyone use throughout the company. It was so complicated that it required a three day training course (a boon to the team of dedicated trainers it required) so that people could understand how to do basic things with the product. Even then, I couldn't make heads or tails of the damn thing. The people who owned the product, however, had been promoted to high positions within the company.

When you have as many people working on as many projects as you do at Microsoft, its critical that you don't have them all reinventing wheels in order to create codebases that they own. That's hard enough when you have a bunch of contrarian developers who are convinced that they would do it better if given half a chance ("Not Invented Here" syndrome is a common disease in the computer industry). It's all but impossible when power and politics get added to the mix.

That's why companies need a core identity. And in my opinion, Microsoft's identity is its traditional ability to turn every product, however inconsequential, into reusable and customizable widgets that can serve as foundation for additional software development.

That goes beyond .NET 3.0 - 3.5, which I think in a few years is going to put real pressure on competing platforms that don't have such a strong API. When I first started working in Internet development back in 1996, I wouldn't give Internet Explorer (then at version 1) the time of day, because Netscape was a more programmable and reusable platform. Within a very short timespan, that all changed, as Microsoft rapidly redefined the browser as a set of components that could be customized and reused apart from the web browser "frame" that constrained the pre-Mozilla Netscape browser. The slam-dunk was when Microsoft managed to put more HTML development features into IE than Netscape (IE 4.0, at the time, was a leader in standards compliance).

The same applies to Office. Traditionally, office automation products were monolithic creatures designed solely for human / user interaction. Microsoft Office wasn't designed that way. Using the same COM API used to offer reusable functionality throughout Windows and across Microsoft product categories, they created a set of automation interfaces that made Office products reusable in ways no one else had bothered to do (at least, not on as large a scale as Microsoft).

Microsoft's past success, in my opinion, is based entirely on its ability to give third parties the tools to create more advanced applications by treating every product as another tool in the development toolbox. This means that more minds than the ones who occupy the halls of Microsoft are working to figure out ways to make Windows a more interesting platform.

Some might argue that consumers have changed. Usability is all that matters today, and the the days of the "killer application" are gone (I'm thinking of you, Harry Bardal).

I don't deny the importance of usability in a software market that has taken a decidedly consumer-oriented turn. Microsoft needs to master the art of good design, particularly in software user interfaces, but also in hardware, if nothing else than as a guide for the wider hardware market that bases itself on Microsoft software.

Even so, I think "killer application" raises the bar too high. Small feature improvements made easier through a robust and improving development platform and tool set are more the norm. "Killer apps" are like meteor strikes, and from a software standpoint, are less common than incremental improvements. Incremental improvements are that much easier when you arm developers with the best tools to make those improvements.

Enterprises could be expected to favor good software APIs, as they build more custom software. That doesn't explain why consumers, however, favored Windows disproportionally at home. It wasn't Apple that put a computer in every home...it was Microsoft. That speaks to the strength of Microsoft's core identity, even if it is not something about which consumers are directly aware.

So, Microsoft's core identity is as maker of platforms, a fact which applies to its traditional operating systems as much as to the applications it creates. What ramifications does that have for Microsoft? I'll discuss that later this week.

Editorial standards