Google Wave, Microsoft and Engineers Running the Asylum

This week marked the official death of Google Wave and Steve Ballmer described the iPad as “just another PC form factor”. These seemingly unrelated events highlight the perils of both driving and not driving a company from an engineering standpoint.

Microsoft, The Meteor is Coming

For those of us who started using computers before the internet, there is a stark difference between now and 10-20 years ago in the perception of Microsoft. In the post-IBM era, Microsoft was the 800 pound gorilla in the room that crushed everything else, even Apple, Wordperfect, Lotus and Borland.

As much as Bill Gates is derided in some quarters he was and is a programmer. It has probably been many years since he wrote a line of code but the fact remains that has has written code and understands the process. He is a technical guy. Case in point: My First BillG Review.

In those days, Microsoft was a lot less bureaucratic. Instead of the 11 or 12 layers of management they have today, I reported to Mike Conte who reported to Chris Graham who reported to Pete Higgins, who reported to Mike Maples, who reported to Bill. About 6 layers from top to bottom. We made fun of companies like General Motors with their eight layers of management or whatever it was.


Bill Gates was amazingly technical. He understood Variants, and COM objects, and IDispatch and why Automation is different than vtables and why this might lead to dual interfaces. He worried about date functions. He didn't meddle in software if he trusted the people who were working on it, but you couldn't bullshit him for a minute because he was a programmer. A real, actual, programmer.

Watching non-programmers trying to run software companies is like watching someone who doesn't know how to surf trying to surf.

Steve Ballmer is no Bill Gates.

Ballmer might be a whiz at making org charts, discussing corporate strategy, managing a division, making M&A deals or whatever but he’s simply not technical. He was hired as a business manager and that’s what he is.

The problem is that Microsoft, in spite of everything it’s tried, is a software product company. If you don’t understand how software is created you have, in my opinion, no business running the company. It would be like me, as a programmer, running an airline.

Ballmer isn’t unique or even in the minority here. My personal experience has been the majority of places I have worked haven’t understood software development nor been driven by engineering rather than the business side. Far more common is that programmers are treated as an unavoidable cost centre for which the perfect management metric has yet to be found.

The problem here is twofold:

  1. Microsoft customers are enterprises and PC OEMs not end users; and
  2. Singularity of purpose: Microsoft exists to sell Windows licenses.

The first point is highlighted in another of this week’s stories, Inside Microsoft's internal IE8 privacy battles. Basically, advertising interests won out over user experience.

Windows is so ridiculously convoluted because Microsoft can’t say no to companies that want ridiculous customizations. This is why you have a horrific mishmash of policies, registry settings, etc that mean it’s typically uneconomic to figure out what’s wrong (if not virtually impossible). It’s easier to simply wipe it and start again.

In Commandos, Infantry, and Police, Jeff Atwood references Robert X. Cringely’s book Accidental Empires, which can be used to categorize companies. Startups are commandos. They are aggressively seeking to gain something with nothing to lose. Successful startups become infantry since the army is now so large that it requires structure and discipline. Truly large companies become police that are only interested in maintaining the status quo.

Microsoft’s sole purpose now is to protect its cash cow of Windows and Office. It’s leadership is probably terrified that something will kill the golden goose. What you fear you create. To quote Princess Leia:

The more you tighten your grip, Tarkin, the more star systems will slip through your fingers.

So this is why Ballmer describes the iPad as just another PC form factor because Microsoft just wants to sell more Windows licenses. Ballmer lacks the technical foundation to understand exactly why he’s running a dinosaur so doesn’t comprehend that the meteor is coming.

Google, Today’s Mammals Are Tomorrow’s Dinosaurs

Google has a similar problem to Microsoft in that it has a couple of core, hugely successful products (being the combined search and advertising business) and everything else has either failed, been lacklustre or is yet to bear fruit.

Of course there are successful ancillary products like GMail, Google Maps, Google Docs and so on. Many of these are completely dominant in their space but all of them merely enhance the search and advertising business or simply don’t generate a considerable amount of revenue.

To quote Google’s Revenue Breakdown Shows It Controls Even More Advertising Than You’d Think:

Google revenues last year were $23.6 billion, 99 percent of which came from advertising, the company said. Google doesn’t break down which of its services earn what, and it still hasn’t disclosed share breakdowns from content on YouTube or mobile devices.

Now I say this not to liken Google’s non-search forays as failures (as Microsoft’s forays beyond Windows and Office are almost entirely failures). Google has a different business model. Google is all about people using the Web. The more they use the Web, the more money Google makes. All their successful non-search products share this theme: they lower the cost of or increase the amount or effectiveness of using the Web.

Google is an unusual company in many respects. Larry and Sergey completely nailed search and will go down in history as tech pioneers for doing so. They have created a company that is largely engineering-driven, which shares a couple of parallels to the Microsoft of old except that early Microsoft was all about embracing the external developer community.

That can enable the company to focus on products rather than attempting to quantify programmer performance (a far more typical scenario) but it can have a downside too, which is the whole point of this post.

Despite it’s massive engineering talent, there are areas where Google has been unable to make any headway, namely in the much-hyped social space. Google Wave’s demise foreshadows the increasingly likely (re)entry into the social space, popularly referred to as Google Me.

But the engineering way of thinking left unchecked can have some serious downsides. Consider an example.


We programmers think differently to “normal” people. While we share a lot in common with practitioners of other scientific and engineering disciplines I think it’s fair to say that part of the programmer psyche is still unique. Many programmers don’t realize this and it’s hard to explain to “outsiders” but I shall try with an example.

Sudoku is a number-placement puzzle where the player tries to figure out how to place numbers in a partly filled grid such that certain rules are followed regarding the allowable repetition patterns of those numbers. Puzzles vary in difficulty based on how few numbers they reveal. A given puzzle has but one solution.

Many people enjoy this as a pastime. When first faced with Sudoku I like many programmers reacted in a very different way. I learnt the rules, devised a program that solved any given puzzle, satisfied myself it was correct and then forgot about Sudoku having never done one of these puzzles by hand, which basically defeats the whole point.

Non-programmers will probably find this bizarre. Programmers will almost certainly nod understanding. The point here is that programming trains you to find general solutions. So you can solve one Sudoku puzzle but what would be the point? There are billions of others unsolved! In terms of time investment, you’re better off solving all of them at once!

Google Wave

Google Wave was presented with much fanfare and hype at Google IO 2009. Spearheaded by the quite brilliant Rasmussens (who brought us Google Maps), Wave was touted as one communication platform to rule them all.

Wave was designed to be low level such that any messaging paradigm could be implemented in terms of Wave. Wave could handle a version of email, Twitteresque micro-blogging, IM, even bug tracking and so on. All of these things can be implemented on top of Google Wave.

Or at least that was the theory.

See the pattern here? Much like the Sudoku problem, Google was seeking a general solution to a set of problems.

So while this may be a great engineering feat, it ignored a fairly basic problem: what is the use case for Google Wave? Who will use it? Why? for what?

Those questions are perhaps unfair because many startup ventures begin with an idea and very little idea of what the end will look like. Or at least that end will change several times (dare I utter the overused “pivot” buzzword?). So you can argue that you create a platform and wait for others to find a use for it.

Clearly they didn’t.

Perhaps a better approach is to ask: what is the pain point? What problem is Google Wave solving?

The programmer in me understand the desire to create one platform that can do everything. It’s an alluring siren but don’t lose sight of the rocks drawing ever closer.

To quote Joel Spolsky from How Microsoft Lost the API War:

By the way, for those of you who follow the arcane but politically-charged world of blog syndication feed formats, you can see the same thing happening over there. RSS became fragmented with several different versions, inaccurate specs and lots of political fighting, and the attempt to clean everything up by creating yet another format called Atom has resulted in several different versions of RSS plus one version of Atom, inaccurate specs and lots of political fighting. When you try to unify two opposing forces by creating a third alternative, you just end up with three opposing forces. You haven't unified anything and you haven't really fixed anything.

To paraphrase, Wave didn’t unite anything. In fact it did nothing other than create yet another way we could communicate.

As much as people complain about email, everyone understands the model. A user who gets on the internet for the first time quickly understand the concept of sending someone “electronic mail”. It parallels something with which they are familiar from real life.

Apple and the Click Wheel

Whatever the future holds for Apple, it (or perhaps more accurately, Steve Jobs) will be remembered as one of the most influential tech figures of the computing era to date. Apple popularized and revolutionized digital content distribution, has a brand synonymous with mobile digital music playback and singlehandedly restructured the telecommunications industry.

The first iPod had a speaker behind the click wheel. Its only function was to make clicking noises to give to give the user auditory feedback. Such a feature makes absolutely zero engineering sense. Can you see Microsoft doing this? I can’t. I can’t even see Google doing it.

There is no other company that completely “gets it” in terms of consumer user experience. No one is close. And before you decry Apple for its lack of Flash, the walled app garden, lack of “freedom” (whatever that means today), etc… nobody cares. That includes me.

The point here is that Apple isn’t run by engineers. Apple is completely focused on product and user experience. Anything can be sacrificed to that end. If Apple were run like Microsoft, the iPad would be running some variant of MacOS X. But it doesn’t. Why? Because it makes no sense to put a full desktop OS based on a pixel-perfect pointer interface (ie the mouse) onto a low-power touchscreen device. Apple isn’t concerned about selling OS licenses. They’re focused on the product.


I’m not against programmers running companies. Not at all. Google is a great example of what can be achieved when you give talented engineers the freedom to innovate. Microsoft suffers I believe because its leader has no understanding of how software developments works and how programmers think.

But both extremes can have a downside. This week it killed Google Wave. It’s food for thought as Google ramps up it’s next stab at social.


Unknown said...

Google wave was massive competition to GMail, I guess. It positioning was questionable to say the least.

Anonymous said...

Hate to buzz-kill, but iOS4 or whatever they're calling it now pretty much is OSX. Not just a variant, but the kernel is the same, the apis are the same... etc.

Eli Bendersky said...

I liked your Sudoku story. It exactly mirrors my experience with Sudoku, but I wasn't sure whether it's the ultimate "geek approach". When I first saw Sudoku, I solved one puzzle and got bored. Then I noticed Sudoku is a constraint-satisfaction problem, so I wrote a program that reduced it to SAT and solved the SAT problem with a generic solver. That was much more fun ;-)

Juan Valdez said...

I'm curious as to why you didn't mention some of the downfalls of a company completely run by designers and marketing(i.e. Apple). Do you feel there are any?

William Shields said...

@Juan: much lime any cult of personality or benevolent dictatorship, the weakness is the single point of failure, the leader. Who knows what will be without Jobs?

steve said...

Jeff Katz; no, actually, iOS4 on the iPhone and iPad is quite different. It's based on UIKit and not AppKit. It's quite a bit more slimmed down than Cocoa.

Both are based on Objective-C, which may be why you see the similarity rather than the large changes behind the scenes to handle embedded computing.

Rahul said...

Anyone claiming iOS is Mac OS X is a great example of what this article talks about: programmer thinking. It's not about which libraries, kernels and APIs the codebase shares. iOS is designed from the ground up to provide a great user experience to people using touch-based mobile devices. Anything else is irrelevant.

Post a Comment