Software

Door Number Three

I'll begin this third and (I promise) last column on IT management with a confession: I have been fired from every job I have ever held. This is certainly not something I set out to do, nor did I even realize it until one day my young and lovely wife mentioned that I had never told her about voluntarily leaving any position. It's not that I've had so many jobs, either. This one and the one before it have kept me going for more than 20 years. But they always seem to end the same way. This one might, too. You can never tell.

Most of the times I have been fired it's because I've been judged to be unmanageable, which is to say I won't shut up. The ultimate reason given is usually something minor. The last time around, for example, I was fired because I didn't transfer the cringely.com domain to my employer. They asked me to do it and I said "no." Had they said, "Transfer the domain or you will be fired," I might have decided differently. But they never said that -- never gave a hint of the consequences -- so I assume the real goal was less to get the domain and more to get rid of me.

The guy who had me fired, Stewart Alsop (maybe you've heard of him), ultimately lost his own job for firing me, at least according to International Data Group Chairman Pat McGovern, who told the story to 300 people once at a DEMO conference.

Back when I was a kid and working at WWST Radio in Wooster, Ohio, I was fired for writing those seven unspeakable words in the middle of a livestock auction report. Another kid had been playing similar tricks on me for weeks, but when I finally retaliated he turned me in. I guess I was a threat to him and didn't know it.

One company hired and fired me three times and another company hired and fired me twice. And somehow in all this I've never received severance or unemployment compensation. I just found another job or it found me.

There's a point to all these firing stories and they actually do relate to IT. I'm typical of a lot of IT types. You know us. We are useful but sometimes a pain in the ass. We have opinions and speak our minds and don't suffer fools at all. We stand up to authority from time to time. Sometimes we're wrong. We get fired a lot and hired a lot, too, because we are generally useful, though dangerous.

What do you do with folks like me if you are a manager? At a traditional newspaper you'd either fire me or make me a columnist. And, sure enough, look where I am. In an IT shop you give me a task to do and let me do it, generally on my own, and never EVER put anyone under me, because I am hopeless as a manager.

But it turns out I'm not so bad as a leader.

Weird, eh?

The last two columns have shown that IT is the Cousin It of American industry. We serve the company but often don't feel part of it. Certainly the value structures and lines of authority that function perfectly well for most of the rest of the company don't work at all well for IT. We're vital but at the same time, well, so different that it's hard to imagine a CEO emerging from the IT ranks. It happens from time to time. Everyone points to John Reed, who rose from IT to CEO of Citicorp, but Reed was an exceptional case. He succeeded because his predecessor, Walter Wriston, had an unusual interest in IT and mentored Reed. Reed succeeded, too, because he didn't really come from IT but from Data Processing, which was more hierarchical. And ultimately he didn't succeed at all, by some measures, because John Reed was fired.

So right now let's just accept that it is very unlikely that, coming from IT, you'll ever become the CEO of your company. That means you are instantly off the traditional management track and have the option of either eventually moving on to some other organization or taking what's behind Door Number Three.

Remember Door Number Three from Let's Make a Deal? It could reveal a sports car or a donkey, but whatever was behind Door Number Three was unlike anything you could imagine.

We need a Door Number Three for IT professionals.

I have a friend of 20 years who is in a key technical role at a very large company. He's too vital to the company to risk losing but too geeky to fit in. He's on the craft (non-management) salary scale, but way higher than he ought to be for having no direct responsibility. All he does, in fact, is from time to time save his company from ruin. And even more rarely, he saves all the rest of us from ruin, too, in ways I am not at liberty to explain. How do you manage such a guy? Where he works they have him report to the CEO. The Big Guy has 5-6 direct reports and one of them -- my friend -- doesn't manage anyone or anything.

THAT'S Door Number Three.

We're in an important transition period not just for IT, but also for business in general. Everything seems to be in flux. And that means the old ways of doing things are changing and ought to. And in this way IT is leading -- or ought to lead -- the way. Later this week I'll be making a dramatic shift and proposing the Cringely Energy/Economic Policy, but first I need to drive home the point that, however different it is from the rest of the company, IT is generally the vanguard for a new corporate culture and whole new ways of doing business for the world.

We're in a mess. The world is screwed up and some of that can be traced to the improper use of IT as a financial weapon. But the people of IT actually present many of the answers we need, because they are living much deeper in technology than other parts of the company or of our society.

Think about it. There has nearly always been a class of eggheads showing us a path toward new business models, whether it was Edison and Firestone, Hewlett and Packard, Noyce and Moore, Gates and Allen, or Brin and Page. It takes in each case a generation to happen, but ultimately we all (and I mean ALL -- everyone in the total organization) come to look like the geeks of the generation before. So let's lean into that, get on with the transition, and get past this place we're in right now where nobody wants to be. Let's consciously embrace the next model that's generally running fitfully right now inside every company, down in the more functional parts of the IT department.

What I mean by this is that times have changed and the world can no longer afford even John Reed's world view with its needs analysis, design, debug, test, rollout strategy -- whether we're talking about a new app or a new marketing campaign. By the time the app (or the campaign) is rolled out, the world changed from HTML to Javascript/SOAP/Ajax (or from financial regulation is bad to financial regulation will save us).

At the heart of this is a concept completely foreign to traditional business -- Open Source. What the open source community has demonstrated is the superiority of a strategy that emphasizes early proof of concept, early release, and frequent releases with features added as needed -- probably totaling 20 percent of the features identified in a needs assessment.

This is the new IT strategy we live with every day -- 80 percent solutions because they are fast, increasingly reliable, and keep the end users in the loop from almost the beginning. All made possible because of an open Internet (at least until Comcast succeeds and enslaves us), easily grasped standards and impressive demonstrations by companies like Amazon, Google, Facebook, and a ton of start-ups. Wall Street back offices figured this out long ago, they just never got their boss's bosses to understand.

Last week's column was a utopian vision that simply requires all the old managers to be reprogrammed or accept a bullet in the head. But it is not at all utopian if applied solely (or initially) to IT, where this stuff actually works pretty well.

IT people are most of the time building fortresses or feeling unappreciated -- often both at the same time. Yet to our discredit, we've done a very poor job of explaining or demonstrating or outright selling our utility to the broader organization. Where are our Geek Appreciation Days? Take a Geek to Lunch? Bring Your Geek to School? Taciturn, we disparage our co-workers for not appreciating us while giving them little obvious reason why they should appreciate us.

That has to change.

Door Number Three isn't just an escape hatch for nerds, it is the way business and culture and civic life will be for most of us a generation further into this information age. We're just leading the way. And if we're leading the way let's embrace that role and become leaders.

If, like me, you are likely to be fired, anyway, there's no real downside to this strategy. Let's give it a try.

Insecure wifi

I was at a Starbucks last week, waiting for my appointment at the DMV.  This was the warning that Firefox 3 showed me, when I attempted to log on to the wifi network there. 

Starbucks and TMobile need to fix their ssl certificates.

Leadership

Last week's column on bad IT management and the strong response from readers that followed show this to be a huge issue. There are WAY too many IT managers who either can't or shouldn't manage technical teams. Last week I maintained that having a firm technology base, or at least the ability and willingness to acquire one, was essential for good managers. While readers got carried away with which technical test is the best, I don't think there is much dispute that there are certain aspects of technical management that are helped by the manager being a code god. But that's far from all there is to the job. So this week I want to go deeper and look at what's really missing in nearly every instance of such bad management, which is leadership.

The distinction between management and leadership is a critical one. Management is -- at its very best -- an exercise in coping while leadership is so much more. Last week's simple idea that the manager ought to at least be able to tell good work from bad is exemplified by Bill Gates, who liked to claim that he could tell good code from across the room and that whatever task a team was facing was something he could code in Visual Basic over a weekend. Both statements are nonsense, of course, but Bill knew he had to talk the talk, making him at least an adequate manager.

Does it make him a leader? I don't think so. But let's not blame Bill for that. Let's blame Charles Simonyi.

Charles is the guy who came up with Microsoft's development process -- an outgrowth of his research at Xerox PARC. I covered this extensively in my book, Accidental Empires, but the short version is that Charles came to advocate a strong program manager as the central controller of any development group. One person made all the decisions and as long as that one person was correct 85 percent of the time, it was better to have a dictatorship than a democracy or even a meritocracy. This was an effective way to extend Bill's will to Microsoft programmers Bill would never even meet. And to Charles' credit the system worked well enough if the dictator was really, really smart and the task at hand wasn't too complex. It was perfect for the 1980s.

But it is far from perfect today and represents one of the fundamental reasons why Windows Vista was so late to market and such a mess when it finally shipped. Vista had plenty of management, but not very much leadership.

When I think of leaders what comes to mind first are political and military leaders. We use the term "leadership" to describe those roles far more than we do for what ought to be similar roles in business or technology. This week former Hewlett-Packard CEO Carly Fiorina said that John McCain, Barack Obama, Sarah Palin, and Joe Biden were all ill-suited to be CEOs of major corporations. However badly the statement went over (Carly supports McCain, by the way), her real point was that there are different skill sets for leaders than managers. And she's right to an extent, but it also says a lot about her own tenure at H-P, which was long on management and short on leadership.

Management is telling people what to do, which is a vital part of any industrial economy. Leadership is figuring out what ought to be done then getting people to do it, which is very different. It is a vital part of any successful post-industrial economy, too, but most managers don't know that.

Let's use the U.S. military involvement in Iraq as an example of the difference between leadership and management. As more books are written and stories come out we can see that there is a lot of arguing that goes on inside the military. Officers are onboard or not. They are proposing various strategies and taking positions generally advocating what's perceived as safest for both the mission they have accepted and the preservation of life among their troops. Eventually someone makes or imposes an order, but even then there is a lot of second-guessing in the military, which has to be ready with Plans B through G just in case they are needed. This is a lot different from the image many of us have of General Patton pointing toward Berlin, imposing a singular view and forcing it through.

In contrast to the military, most businesses do a lot less explaining and pondering and a lot more laying down edicts. That's management, which works fine on an assembly line, but not at all well building a big software application or winning a war.

Janna Raye is someone I worked with at InfoWorld half a lifetime ago who has built a consulting career on understanding this stuff and helping companies to transcend 20th century corporate hierarchies and become what she calls "fractal organizations." Janna's consulting business is called Strategems and here is her take on this issue:

"Modern corporations suffer from systemic-level issues that emerge in top-down hierarchies. Managers are there to control staff and budgets, not to lead. Although you can make valiant and often successful attempts to control things and processes, you will never again be able to control people. We've evolved, basically, and the information age has had a lot to do with it. So we still "manage" companies the same way as when we actually operated assembly lines in America--the good old days! Now, people need leaders, not managers, and that's what a fractal organization enables.

"In fractal organizations, it's the staff deciding how to continuously improve processes in their functional areas for efficiency of time and resources. These organizations thrive with a new pay model also, based upon results or value of work delivered and not how much time it takes to do the task. Those who are really good will get to go home early! These are not the organizations that are shrinking. Like galaxies, they continue to expand, actually aided by a strong gravitational pull of the leaders at the center. Those who do it well create a compelling vision and keep it alive. They allocate resources to projects that align with the vision, and reward arm- and team-cluster leaders for the creative ideas their staff bring to the organization. It's a shared vision and collective goals that are missing from the vast majority of organizations, which is why failing projects continue to drain resources. Really caring about what you do and feeling proud to be a part of something special and wonderful is what every human desires, even if they say they don't."

So what's Janna's model for the ideal 21st century organization? Pixar.

"They let creativity run wild at Pixar!" says Janna. "Ed Catmull, Pixar's president, wrote an article for the Harvard Business Review on their collective creativity. Ed and John Lasseter (and sometimes Steve Jobs) are in the center of the galaxy, keeping the gravitational pull strong and the company rotating, so to speak. Around them are the directors, who in the fractal org model lead the arms (no more divisions!). Each film team or cluster, from the storyboard artists to the renderers, goes through the iterations necessary to achieve the best results. Everyone is on board with creating exemplary films and they are relentless in demanding only the best. But in this process, as they say, they have to get all the "sucky" ideas out first. If they don't reveal all the ideas, they don't get all the perspectives and therefore might miss something important. In quantum physics and information theory, this relates to the observer effect and the importance of acknowledging the perspective influence of everyone in a scene.

"In top-down hierarchies, the opposite occurs, yet the people on the front lines are the ones dealing with the evolutionary changes going on around them and are the best source of ideas for solving tactical issues. You wouldn't need "change management" if you made continuous improvements at the functional level the responsibility of every individual and team cluster. Yet Pixar makes incredible films in this manner, so you could certainly accomplish it anywhere, even in a supermarket! In fact, some of Whole Foods' leadership practices are fractal -- in-store teams make decisions about products and their placement, based on their observations of customer patterns.

"Most start-ups are fractal in their nature, especially those that have exciting visions and get everyone on the same page with collective purpose, goals, and objectives. Most investors, however, are bought into the conventional org chart; when the company devolves into top-down, the turnover begins. That's because of the internal competition that emerges in top-down organizations. The perception is that there's only so much room at the top. At each level of management, the competition increases as cooperation decreases. Thus are created the ubiquitous "silos" of information that thwart collaboration and encourage redundant, wasteful business practices.

"Managers are supposedly promoted because of their ability to outperform others and not because of an intention to provide inspiration, guidance, and mentoring to their staff, nor are they openly rewarded for this behavior, even though it usually produces a healthier bottom line. The usual way of rewarding based upon meeting financial goals and managing budgets keeps the focus on short-term financial results only, whereas continuous improvement leadership by frontline staff creates more long-term successes.

"When managers don't mentor staff, focusing only upon numbers and bossing people around, it leads to an illusion of control, of which there's no such thing. In these situations, they begin to feel they must continually prove their worthiness and so defend their territories against possibly brilliant staff working "beneath" them. This is a systemic issue, not a personality quirk, though some personalities are more susceptible than others. In most companies, the idea of climbing over others on your way to the top and throwing people who get in your way under buses is de rigueur. The top-down hierarchy was designed to manage industrial-age processes, not information-age challenges. You didn't want the door guy getting creative when attaching the door. Nor did he need to collaborate with the bumper dude. The information age is vastly different. Each scene we're in presents new circumstances and opportunities.

"Pixar claims they have a meritocracy. This is a good description of the atmosphere that emerges in fractal organizations. Google was likely more fractal in the beginning, before they brought in managers trained in top-down hierarchies and engrained with the accompanying behavioral patterns, such as knee-jerk resistance to ideas they haven't thought of themselves. Not everyone is like this, of course, but usually those who aren't have had confident mentors themselves who encouraged creative participation. In Catmull's HBR article, he says they insist that everyone in the company contribute ideas, across all functions and levels, or they will perish. Interestingly, he mentions the difficulty in getting new hires to feel comfortable with this process. This results from cultural-level systems that keep people in competition rather than cooperation. Though Catmull tells readers what they do at Pixar and why, he doesn't instruct on how to make the organizational changes that enable this approach to creativity."

I guess that last part is Janna's job.

All this fractal stuff is interesting, but I'm guessing it is also difficult to implement, because it may describe Pixar but it DOESN'T accurately describe Apple -- the other place where Steve Jobs is king. Still, Apple's product success shows that something is being transferred from one company to the other. It's just that at Apple, Steve Jobs can't make himself stay out of the way.

Fire Your Boss

This week marks the seventh anniversary of the 9/11 terrorist attacks in the U.S. This week is also a time when the world economy is under stress comparable or greater to that imposed seven years ago. Whatever you are feeling in your wallet, I can't overemphasize the impact the current global credit crunch is having on our economy and that of other nations, including Germany and Japan. We're in a mess -- one that is at least TWO YEARS from being resolved no matter who is the President. This matters to a technology columnist because it is something the technology community will eventually have to address, just as we did 9/11. I think some coping skills are in order.

First let's take a look at a small part of my column from September 13, 2001 -- a column that wasn't especially popular with readers at the time, but I think stands up pretty well with time:

" 'To a man with a hammer, everything looks like a nail,' wrote Mark Twain. In the current context this means that the organizations charged with reacting to this catastrophe will do so by doing what they have always done, only more of it. Congress, which controls the budget and passes laws, will want to pass laws and to allocate more money, lots of money, forgetting completely about any campaign promises. The military, which is the nation's enforcer, will want to use force, if only they can find a foe. The intelligence community, which gathers information, will want to be even more energetic in that gathering, no matter what the cost to the privacy of the millions of us who aren't thinking of terrorist acts. And agencies like the Federal Aviation Administration, which regulate, will want to create more stringent regulations. Now here is an important point to be remembered: All these parties will want to do these things WHETHER THEY ARE WARRANTED OR USEFUL OR NOT."

In 2008 we're facing cuts in IT that are prompted by economic decline. Many of the IT shops I talk to are in denial about this. Many more, while not in denial, are making bad decisions. I think this is a good opportunity to do some housecleaning that probably should have been done years ago. If you have to cut your budget by 10 percent, where do you cut? What if you have to cut by 30 percent?

As I have written before, one of the great problems in IT management is that the big bosses typically haven't a clue what is happening, what is needed to happen, and what it all should cost. There is a role for trust here, but if the Big Guy is signing off on a budget he can't even read, much less understand, well something is wrong. Some IT departments like this, of course, just like my students liked it when class had to be cancelled (they liked getting LESS for their money), but in tough times, facing reality and speaking the truth is usually the best course.

Because power in IT organizations tends to be based on head count, preserving jobs takes a priority. And when jobs have to be eliminated, they tend to come off the bottom of the organization when they should more logically come off the top -- or at least from near the top. A tech who directly helps users is more important than a manager who can't manage. This is especially true if that manager is making 2-3 times as much as the tech.

If your boss doesn't understand your job enough to describe it in technical detail, that boss is in the wrong job.

If you are managing an IT shop and can't write the code to render "hello world" in C, html, php, and pull "hello world" from a MySQL database using a perl script, then YOU are in the wrong job.

I should point out that these latter tasks can be copied and pasted straight from properly composed Google queries. They aren't a test of programming knowledge at all, just of the ability to use the Internet. Yet many technical managers will fail and should get the boot as a result. You can't manage what you can't understand.

Think about whole projects that can be chopped, too. What's really needed, after all? That knowledge is in your organization, though often not where it is available to the decision makers. The essence of efficiency is doing only the parts that are absolutely needed and almost every shop has at least one project that everyone except the big boss knows is either pointless or hopeless. Cut it.

One can argue, of course, that MORE IT, not less, is in order, and maybe that argument will work. But make it only if it is true.

I think a good argument can be made for embracing cloud computing, even to the extent of eliminating data centers and facilities. We're very close to the point where relatively few organizations really ought to have their own data centers.

This could also be a good time to embrace open source tools. Yes, there is a learning curve, but the price is right and I can argue that open source quality is substantially better.

Oh, and cancel those contracts with Gartner, Forrester, IDC, etc. You'll feel better in the morning.

Some folks who WON'T feel better in the morning are the next class of IBM employees to see their jobs moved overseas. I understand that there is a new round of cuts coming in October and it will be different from the "death by a thousand cuts" that has been happening for the past year. The technique is the same, of course -- cutting unneeded workers in the U.S. while suddenly needing virtually identical (if younger and cheaper) workers in India and Argentina. It's pure coincidence. Yeah, right.

If you were disappointed with Apple's product announcements this week, take cheer from knowing that more announcements are coming, including new MacBooks and iMacs. You don't have to take my word for this -- just look at the closeouts Apple is offering on current models. Christmas is still the most important quarter for Apple so they won't let these announcements wait too long. I just wonder how they slipped out of this week's event.

And finally, I am surprised to admit that the latest version of 64-bit Windows Vista seems to be running pretty darned well on my desktop. No driver problems, 32- and 64-bit apps seem to be running well -- why hasn't Microsoft been shouting about this? 32-bit Vista still sucks, of course.

links for September 7, 2008

• Technical Brief for Windows Home Server Media Sharing