Playing the Odds

How many companies hire people in the hopes that they will do something great? More specifically, how many look at the odds of someone doing something great and say "let's hire more people, to increase our chances of that happening?"

This is the modus operandi of a hit-driven business, where you play the odds hoping for the hit that will cover all your expenditures and make oodles of cash on top of that.

The problem with hit driven businesses is that they treat people like a pull on the slot machine rather than an investment. They assume that hits come as bolts from the blue, rather than the result of long years of effort and preparation. The magical hit is a much more exciting prospect, so we use our built-in filtering mechanism to remove data that suggests otherwise; thus our fondness for luck. If you don't produce a hit, and quickly, then they're off looking for the next possibility.

Unfortunately, it seems that this kind of short-term thinking is predominantly what business schools teach.

Why Do We Retire?

There was a dinner party for an author friend of mine, hosted by a couple here in town. One of the hosts, Bob, told the story of retiring early -- he had done quite well and perhaps gotten tired of the business world. At that point he got a pained expression on his face and you could tell he was remembering something deeply unpleasant. He said, simply, "you need a plan."

Apparently it was quite common for men of my father's generation to die within a year of retiring. So common that big companies in Detroit and elsewhere factored it into their retirement plans: you can offer very generous retirement benefits if people are not around for very long. Apparently the identification with your job is so strong that men of that generation essentially felt useless once they retired and dropped like flies, almost as if there's a genetic switch thrown when you are no longer contributing.

In his new book Cognitive Surplus, Clay Shirky, like many of the other authors I've read recently, makes the point that we need to feel autonomous and competent. I believe we also need to feel useful, and for men of my father's generation, "useful" was rigidly tied to "job."

I just spoke to someone who is very excited about retiring in a year, because then she can go and do something useful. The aforementioned Bob ended up volunteering on the hospital board (although he still seems to be looking for more to do). I've never been interested in retiring. Why can't I just continue to enjoy life and be useful? Why, especially, do we think of our "working years" as a time to put in the hours to make the money so we can retire and do the things we really want to do? And think of all that lost experience and productivity, that a business could benefit from if it makes itself a place where people want to be.

Repeated Mistakes

(Excerpted from Disruptive Television, originally published March 14, 2007)

It's amazing how many of the really basic business mistakes get repeated over and over. One classic is changing the name of your business, which is almost universally disastrous. Then there's the inevitable "moving the corporate headquarters" that always seems to happen when a new CEO is hired. A big study is undertaken, and the results of that study firmly show that the best place for the new headquarters is whereever the new CEO happens to live. Imagine that. But the worst mistake a business can make is to stop thinking about what serves the customer. When that business starts to say "what serves me best?" or worse, "what's the quickest way I can make the quarterly profit sheet look good?" then they have lost their way. Fortunately, in the age of the Internet, the competitors are usually quick to step up and fill the void.

The Distrust Effect

Tell me some good news. This is an only-slightly-veiled directive that tells you to lie, a little or a lot. It's explicit here, but implicit everywhere in the org chart. As information travels up a hierarchy, it distorts and becomes "happier," because everyone wants to tell their boss good news. This is called the mum effect, and it means that cluelessness increases the higher you go in the hierarchy.

But that only describes what happens on the way up. What happens when information travels downward in a hierarchy?

Everyone knows their supervisor is slightly clueless because (1) they made the supervisor that way, by giving only good-ish news and (2) they've seen the supervisor in action, making decisions based on their slightly-distorted world view.

Supervisors know that subordinates often resist directives. In Hard Facts, the authors even say that ".. a case can be made that when leaders are wrong -- and people don't have the power to reverse their commands -- that ignoring orders, delaying action, or implementing programs incompletely may be best for all involved." Subordinates will, on their own, sabotage the supervisor's directives! (There's economic shirking in action!).

As information flows downward, the hierarchy appears to produce distrust, in both directions. Perhaps we should call it The Distrust Effect.

How do we fix this? As long as an organization is about power (and extrinsic motivators), then the signals to employees will be around serving that power. Look at any organization: the more time people spend serving power, the less they spend serving the customer. It takes someone special to "speak truth to power," another data point showing that information flowing to power becomes corrupt.

We are trained from birth to pay attention to, respect, and gain power. Early on you learn that rules are more important than people, fourth graders are better than third graders, and leaders are superior to those they lead. That pattern is not something we are going to shake off easily, not without a system that pushes us in the opposite direction. One option is to create a new organization that isn't about power, but perhaps there are techniques to help take power out of existing organizations. I'm skeptical, but I am looking for such techniques.

Objective

To create a generic model for a self-organizing, self-managing, auto-stabilizing business that maximizes positive evolution by changing from the bottom up, and minimizes organizational costs.

Imagine me, some years ago. I've managed to temporarily avoid joining the real world by going to graduate school, but that's at an end. I'm facing the prospect of entering corporate America, but just in time I receive a time capsule from myself in the future, containing the plan for this business. I show it to some friends, and we decide we want to continue living like poor graduate students -- we're already comfortable doing that -- and use this model to start a business.

The business model allows us to skip the usual struggles, fights and hardships of the typical startup. When the company hits the points where traditional companies have upheavals from growing pains (because they are changing from one way of organizing to a different one, on their way to becoming a typical dreadful corporation), we don't even notice it -- because our fundamental principles are still working just fine. In fact, we never face the prospect of having to become a different company because of our size, or any of the other usual pressures that force a startup to lose itself.

In the past, the idea of starting my own multi-person company has just seemed like so much hassle. It has never appealed to me. Similarly, my experiences helping to organize conferences like the Software Development Conference has not made creating one of those appealing, either ... but when I started doing Open Spaces Conferences, everything changed and creating conferences is easy, a delight, and I do it whenever I get the chance. I want to find the formula that makes creating and growing a business as delightful and effortless as creating an open spaces conference. In short, I want to create a business model that makes it nearly irresistible for me to start a business.

It's possible this business model could work for existing companies but I suspect, at first anyway, that it will only be in the domain of startups. It will be too different, too many changes for a traditional company.

What artifacts will this effort create? First, this web site, which seems to be evolving into my research notes and first drafts. I hesitate to write a book simply because that's what I already know how to do (is it a knee-jerk response or the rational path?), but books can be useful and I do know how to write them. The business model itself, of course, but that needs to be as brief and simple as possible, almost a manifesto-like document similar to The Agile Manifesto and its Principles. Other artifacts might include conferences, videos, and some model for viral teaching events (once someone sees how it works, they can go create their own teaching event, much the way it happens with open spaces).

(Because I get this question a lot). I don't yet know how I'm going to make money at this, and the central goal is the new business model, not how I earn money from it. I believe I will eventually be able to support my efforts, through something -- as yet unknown -- that will emerge during this development process. But I don't expect that to start happening for a couple of years.

Distributed Decision Making

Now that I have flattened the hierarchy, how do we make decisions?

A disclaimer: I rely on a combination of "flash of inspiration" and "let's do the opposite of what isn't working" for some of these ideas, including this one. Or you could just say that I pull them from my nether regions.

Premise 1: Everyone in the organization is equal. Of course, we don't really know their actual skills, strengths and abilities, and what they might contribute to the organization, given the opportunity. However, traditional hierarchies only pretend to know these things, and by slotting people into "positions" they prevent those possibilities from emerging. So the closest thing we can do is just assume everyone is equal, pay them equally, and create an environment to draw out their greatest abilities.

Premise 2: Without knowing specifics about a particular decision, we can say, approximately, that no one is more qualified to make a decision than anyone else. This is wrong in any particular case -- obviously, once you know what the question is, there is going to be an optimal person within the company to make that decision. What is the probability that this person is part of your senior staff? The traditional hierarchy says 100%, which is clearly wrong, and the quantity of bad executive decisions demonstrates this. On the other hand, we just can't know who will be the exact right person for a particular decision. So we'll make a rough starting approximation and say that everyone is equally qualified, that the decision will "find it's own level" and that The Wisdom of Crowds will correct mistakes.

Premise 3: A culture of experimentation is one of the most critical pieces for a company's forward progress. There must be no inertia involved in saying "let's try that!" The culture must support experimentation and an environment of "ask forgiveness, not permission."

Premise 4: Rapid decision-making and easy iteration to adjust those decisions produces an agile company.

Based on these premises, here's how distributed decision making works:

• The first person who encounters the question makes the decision

We're all in this together and everyone is equal, so each member of the organization keeps the organization's best interests in mind. Thus, any member of the organization is likely to make a good decision.

• All decisions are recorded transparently

If you make the decision, you record it. Ideally, this is a system like StackOverflow uses, so it's very easy to find existing decisions (to keep you from making the same decision over and over). When the decision is recorded, it also goes on a list for review.

• Anyone can question a decision

That person then becomes responsible for reviewing it inside the recording system, and bringing it up for re-evaluation.

• If you can't make a decision, you can delay it

You might, for example, know that someone else would be more qualified. In this case, you still record the issue but it goes on a "pending decision" list.

• If you can't make a decision and you can't move forward without it, press the "stop" button

This is just like pressing the stop button in a Japanese car assembly line -- you don't want to do it very often, but if it's holding things up, then it's a decision we need to stop and address. The stop button will happen less often as an organization matures.

Decisions made by the group use some kind of majority of those voting (not everyone will be available all the time). Remember, this is not an exercise in power nor are any decisions cast in stone. The act of voting just attempts to get the wisdom of a group, so the decision can be made and we can move forward. And if that decision doesn't work, we can revisit it.

If a person makes a decision that already exists in the record, then that person is responsible for reconciling the two recordings -- not as a punishment, but because that person is closest to the problem (they were unable to find the existing decision so they have the best feel for what's wrong with it).

I know this sounds pretty strange at first. Give it some time to sink in, and keep in mind that we're not trying to make something perfect, because it never can be (indeed, belief in perfect control is probably what makes most organizations the way they are). We're trying to create the loosest, fastest decision-making system which gives people the ability to easily make most decisions, and trusting that they can work out the special cases.

This system is itself not cast in stone; it can be reviewed and changed. Try, however, to remember the principles of "simplicity" and "trust" when doing so.

The Destabilizing Effect of Power

In physics, power is force times velocity. Power isn't happening unless it's pushing something around.

The analogy mostly holds for power in a hierarchy -- it must be moving something to be expressed.

Systems of people want to be stable -- the people within that system make efforts to maintain stability.

So, you have a system that wants to be stable, and power that needs to push something. There's your battle.

Maybe the Worst Thing

You invest in something or someone, and, as time passes, you get a sneaking suspicion that it might have been a mistake. The thing begins to veer away from what it was supposed to do, or the person starts dragging their feet on the projects they had seemed enthusiastic about during the interview. They haven't exactly stopped working, but if you draw a productivity chart you can see it will eventually cross zero. The trend is not promising.

Then there is the time you spent researching and buying the thing, or finding and interviewing the person. Add to that the time integrating the thing into your system, or training the person (and paying them to learn). All this will be lost, and then you imagine the additional time to get yet another person or thing, and it's all terribly daunting and schedule-blowing. It sure would be nice if the thing started working right, or at least if it would stop working quite so wrong. The person might change, too. It's been known to happen. Things could get better all by themselves.

We are terribly loss-averse, it seems. We'll do all kinds of stupid things to avoid loss, missing out on tremendous opportunities in the process. My own theory is that this came from natural selection: take too many risks and you starve. Or you're scouting for new hunting and gathering ground, and you've gone quite far in a new area. You've invested a lot of effort getting there. Now, you think, maybe just over that mountain I'll find something. It probably worked really well up until the last few thousand years.

Experiments were risky and expensive, so making mistakes could be dire. But now, in most developed countries, experimentation is relatively safe and cheap. You can afford to buy printers until you get the right one, but doesn't it just grind you when you buy the wrong one? You feel stupid, used, tricked, a failure. It's probably encoded in our genes somehow: Don't screw up! You shall feel pain if you do! Plus, if you just had the right information, you'd get it right the first time.

It's more probable that this pattern will cost you money, sometimes a lot of money. Following the "sunk cost" approach (we bought it/hired them, we've committed, we've got to use it/them) is unbelievably expensive. It's absolutely illogical and unscientific: you keep getting evidence that a path is wrong (and, knowing what you know, you wouldn't choose that path now), but you keep doing the same thing, hoping the results will be different (this is one definition of insanity).

We are so averse to making mistakes that it might very well be hard-wired. And yet, success in the modern world comes through experimentation. The better you get at experimentation -- and I'm not talking formal, scientific-method stuff here, I just mean get ideas, try them out -- the more interesting, unusual, creative things you will make, and that's what drives the economy: the potential energy and entropy flow of ideas from "unique" to "commodity."

I think the only way we can become more experimental is to change our fundamental belief system, mostly around failure. Changing a belief like this won't happen quickly or easily, so in the meantime we need a device to, in essence, trick ourselves into doing the right thing.

A variation of this device should allow us to face the problem of dealing with an association that isn't working out, or isn't trending positive. Perhaps we can draw from the Netflix employee evaluation technique, or some of the Getting Things Done techniques, or research from a book like Switch.

Growing Pains

All traditional businesses, as they get larger, go through "growing pains," but I think the company actually changes fundamentally, transforming itself into a different creature. At these points there is often an exodus by people who know better and don't like the new culture and rules: someone who is with the company as a startup may not find the change in culture as the company "matures" to be pleasant or interesting.

Here is an excellent example, where a new CFO of a 3.5 year-old startup company decides that free sodas are costing too much, triggering an exodus of the best talent. This should produce panic, but instead typically involves an adjustment of the world view further towards the "new" version of the company, where the message is now: "The company is more important than the employees."

Hewlett-Packard made strides in solving this problem by establishing size limits on a particular group, so if it went over a particular number (something like 1200, I think, but I don't remember), they would break the group into two. That way it stayed a kind of "small town" where you knew everybody.

My hope is to create a business model that doesn't have to change its culture and nature as the number of participants changes.

Embracing Fallibility

This article from the Boston Globe makes the case that most decisions are made through inductive reasoning -- a.k.a. "educated guessing," and those decisions are generally quick and very often right. Inductive reasoning seems to produce a kind of optimal process and probably came from evolutionary selection: those who sat and pondered trying to get the exact right solution got eaten, while those who went for the fast solution that is usually right did not.

We easily get stuck in a backwater of induction. We can form irrational beliefs based on induction and, without training, we can't work out how to escape. Management theories seem to be dominated by these irrational beliefs, probably because we put so much faith and power in the hands of our leaders. Those leaders promise "rightness" and being "the decider." They don't want to be wrong, and we don't want them to be wrong, either. Inductive reasoning by itself creates a culture of infallibility. Mistakes are hidden -- we don't like the way it feels to be wrong. So mistakes accumulate, eventually causing much bigger disasters than if the mistakes were acknowledged right away and corrections made.

This is the pitfall of leadership, the inevitable cycle of (1) "magical rightness," (2) hiding mistakes and (3) eventual downfall. People in the company know when it's happening. They can see it the moment it starts, the point when a leader becomes disconnected with reality and believes his or her own PR.

It's hard to know whether it's our culture or our nature that makes us so deeply hate the act of being wrong (I suspect culture, myself). But it's currently ingrained within us and will take some time to grow out of. In the meantime, putting all power in the hands of any of us is crazy, because if we hate to be wrong and we hide it, we can't fix our mistakes.

To repair the problem we (1) Create a company culture that embraces mistakes (the Boston Globe article was written by the author of a book that promises to show how to do this) (2) Create a decision-making process that leverages The Wisdom of Crowds, which I shall attempt in a future post.

Executive Compensation and Moral Hazard

Guest post by Dave Bartlett.

I did a Master's thesis (you can find the paper here, without appendices or title page) on executive compensation back in the mid-'90s with data from the late '80s and early '90s.  The basic hypothesis of the thesis was modeled around some of the topics you are exploring here.  Basically, the work in this thesis looked at data for public corporations to see if there was any correlation between the structure of executive pay and the return to shareholders of the company, this is the "employee - employer moral hazard dilemma."

It is your typical academic paper with lots of formulas and data, throwing around term after term.  I did this project because I was intrigued by a CEO who in the '80s and '90s had difficulties while running two high profile companies Bendix and Morrison-Knudsen.  He effectively sold high value assets for short-term cash flow at the expense of the long-term success.  Essentially, he would record results for a couple of years then the cash flow would collapse and he would show record losses.  After years at the helm of both companies he was asked to leave.  How could someone do this to one company then be asked to do it again at another company?  That's what got me going on this topic.

As stated in the introduction: "Moral hazard manifests itself in many ways but most frequently it bubbles to the surface when decisions are made.  The office of a chief-executive-officer (CEO) is the single most important focus in analyzing how decisions in an organization are developed.  A morally hazardous decision will be introduced into an organization when the objective and goals of the CEO or any decision-maker are not in alignment with the objective and goals of the owners or shareholders of the organization."

The hypothesis this research tests revolves around executive compensation structures that reflect different characteristics influencing the ability to monitor the corporate agents in an effort to reduce moral hazard.  The paper investigates a sample of CEOs whose incentive compensation is based on a combination of salary, bonus and stock-option awards.

This paper made three contributions to existing research at the time.  First, it models the relationship of moral hazard between the agent (CEOs) and the principal (shareholders or owners).  Second, it identifies mechanisms that mitigate moral hazard problems that can be common in organizations.  Third, the results add to evidence in the business literature on executive compensation management with respect to moral hazard.  Previous research in this area has focused on moral hazard in franchise contracting and suggests that agents have incentives to align their goals with those of the principal and reduce moral hazard in the decision making process.

In the end the results of the regressions did not support the first hypothesis postulating that larger corporate size will show a stronger emphasis toward an overall CEO compensation weighted heavily toward stock and stock options.  It is possible that other influences were operating on our formula.  CEOs operating in smaller organizations may take stock and stock options because the share price has more growth potential than in a large mature organization that is fully capitalized and can afford to pay large salaries and bonuses to CEO’s in good standing and with strong track records.

The second hypothesis that organizations with risky cash flows would want to align CEO’s compensation to stock and stock options had mixed results.  There seemed to be a correlation between the cash flow variance variables and the cash flow horizon variable.  Which makes sense if you have an organization with risky cash flow (industries like disaster insurance or mineral exploration) or a long horizon (think pharmaceutical industry) you would want decisions made balance current and future success of those cash flows.

The results did support the third and fourth hypotheses.  Organizations with longer cash flow horizons did seem to link their executive compensation to corporate share and stock options.  In our fourth hypothesis we stated that organizations in regulated industries would not be inclined to link pay to corporate shares and stock options.  These industries are in effect being monitored for the shareholders by an outside organization.  This was indeed the case with all coefficients being negative and most showing strong significance.

I walked away from this exercise with a couple of non-academic insights.  First, I don't feel being a public traded company should be a priority for most companies.  The loss of organization control and the focus on short-term return seems only worth it in the extreme.  I can't quantify that extreme but it would be a decision akin jumping off a really high cliff -- it's hard to know exactly what will happen but chances are some else will wind up cleaning the mess.   This was during the time of the Internet explosion and everybody and their sister was starting a company that was going to go public.  Secondly, most of this thesis is about the finance and accounting side of corporations and that is what business schools teach.  In reality most companies and the people working in them focus their energies toward the completion of some project that will produce some product or service.  Organizations, their leaders and strategies should be focused on those tasks and not the formulas and metrics this paper analyzed.  These formulas and metrics need attention but not complete organizational focus.

Stamp It Flat

Two fundamental characteristics in the model of Economic Man are that people are lazy and don't want to work (economists call it shirking), and that we will steal, given the opportunity. If you believe these things, you will most likely create a management system that looks similar to most of the ones we see today: designed to prevent employees from shirking and stealing.

In Drive, we learned that people actually want to be productive. What if people don't actually want to steal, either? (Netflix tells its people to act in the best interests of the company, dispensing with numerous rules and accountants in the process). What if these core principles of last-century management are actually wrong?

We won't go so far as to assume that some people don't behave antisocially. The traditional hierarchical system is designed around those people, and in many ways has become designed for those people. Instead, we shall create a system that is uninteresting to those people, even to the point of invisibility or repellence. They look at this system and see no way to gain power and control over others, so the organization is completely unattractive to those people; they cannot imagine why anyone would want to work there, and they look elsewhere.

Suppose we can assume that people -- or at least our people, the people that end up at our company -- will be productive and they won't steal. What does a business look like if you can believe those things?

This system doesn't try to fix people by forcing them to behave in appropriate ways. Nor does it try to fight the tendencies of the traditional hierarchical system, because the pressures of the hierarchy will always reassert themselves -- you can build a company that is temporarily different, but when you expose it to "normal" corporate pressures (by hiring a "professional" executive team, for example, or by going public), those pressures turn it into the same undesirable result.

Instead, it changes the foundational rules so that the organization (1) attracts the right people and repels the wrong ones, and (2) grows and evolves without changing its fundamental character (avoiding the "growing pains" points of traditional companies, where they basically change who they are).

The basic principles for this kind of organization should be few: We have trouble remembering more than a few things, and it's very important that we keep these principles clearly in mind at all times (the Rider needs clear direction). In addition, the same principles that make this organization attractive to people who have positive social behavior should make this organization repellent to those who behave antisocially; in particular, those who desire power and control over others.

In A Modest Proposal, I suggested a flat pay system, because I find that to be one of the most powerful signals that We're All In This Together. Chapter 8 of Hard Facts is filled with evidence of the corrupting effect of power within corporations, and arguments against the cult of leadership. Unfortunately, the best suggestion they could make was to say, in effect, "Don't let it go to your head" (if you find yourself in a position of power). The alternative was too radical to consider: take power off the table. The answer lies not only in the pay system -- perhaps the answer is to take the whole hierarchy and

Stamp It Flat

In Here Comes Everybody: The Power of Organizing Without Organizations, Clay Shirky makes a compelling case that the hierarchy was created for communication. Even if that communication is flawed, as we see in the mum effect (information gets "happier" as it goes up the hierarchy), enough information flowed for organizations to stumble along. But, Shirky says, we now have electronic communication which can replace and flatten much of the hierarchy. What if, with some adjustments to our system, we can flatten it completely?

This removes power from the equation. Nobody is "better" than anyone else, and nobody controls anyone else. Nobody gets to help themselves to perks because they have the power to do so, or abuse others because they feel like it. Corporate psychopaths will look at such an organization and not only see no advantage for themselves, but will see a distinct possibility of being exposed. There is no need for a No Asshole Rule, because those people will not be interested in joining the organization in the first place (indeed, they will flee from it).

I didn't invent idea of a leaderless organization; in 2008, The Starfish and the Spider: The Unstoppable Power of Leaderless Organizations was published (see here for a nice summary), and I suspect there are others who have suggested the same thing.

If we replace leaders, we need to replace their responsibilities. Leaders:

• Tell us what to do. Let's assume we can figure out what to do.
• Make sure we don't shirk. Let's assume we want to be productive and useful.
• Monitor and evaluate our output so they can determine raises. Both tasks are eliminated by a flat pay system.
• Communicate with higher-level managers and express their desires to us. (And incidentally filter it both ways). We use electronic communication with transparency, so filtering is not necessary.
• Prevent Mistakes. How often does this actually happen? How many mistakes have been caused by managers, via the aforementioned filtering, and through over-managing?
• Make decisions. This one can't be so easily dispensed with -- someone with authority must make decisions, right?

Well, let's think about that last one. Is it reliable to assume that managers always know more than those they manage, and can make the best decisions? How often do managers make the best decisions, vs. just flipping a coin -- or even letting subordinates make the decision. And once a manager is invested in a decision, how difficult is it to change when it becomes clear it was wrong? We've learned that groups make better decisions than individuals via crowdsourcing. Then there's the Dunning-Kruger effect, making managers believe they are smarter than they actually are.

Might it be possible that in many cases, just making a decision and seeing what happens -- correcting later if necessary -- might actually be far more effective than relying on the traditional hierarchy to do it?

I will suggest a scheme for distributed decision making in a future post.

Evidence-Based Management

(This is not what I promised in the previous post. Still working on that.)

The final chapter (9) of Hard Facts is titled Profiting from Evidence-Based Management, wherein they attempt to summarize and focus the results of the book, and give you a system to practice, while acknowledging that many efforts are prone to failure.

I love the idea of evidence-based management. It says, "Let's apply scientific thinking to management" -- ironically, the very thing that Frederick Winslow Taylor claimed to do in The Principles of Scientific Management, published 99 years ago, while at the same time he was fabricating his data and results. The core belief that Taylor begat and propagated was that management is a science. Thus, managers could have the same control that scientists have over their experiments, and management became a process of tweaking knobs in order to optimize output.

The irony is not simply that management is not a science, but that the business schools, founded in the wake of Taylor's "discovery," and all the students that they manufactured, learned so well from the way that Taylor worked (inventing data and conclusions to produce any desired outcome) that they never learned what science actually is, and went on to fabricate fantasy realms with their own "theories," and models based on those theories. Nothing had to actually work -- it just had to be a good enough story to be salable. The business of management is the invention of new theories of management.

That's why we need Pfeffer and Sutton to come along and say, "Make decisions based on evidence!" This is kind of dumb, when you think about it: After 100 years of management that claims to be scientific, we have to explain that "science" means "based on evidence."

Actually, science means more than that; it means we can create a consistent model (which means reproducible results, not sounds good to me) that will accurately predict behavior. And that is something that management won't ever be able to do -- it can make educated guesses about trends and tendencies, but predicting the things that it wants and claims to predict isn't going to happen.

Management is not a science but, as Hard Facts points out, it is possible to get help from science to improve management -- and that is what evidence-based management is about. I like that the phrase puts it in the correct perspective: not science-based or even fact-based, because that promises too much. Evidence-based says: "We are going to pay attention to things that actually happen, not just things we make up in our heads." It says that management is an art, but one that can be helped and influenced by what we learn from science. We can't pretend we are in control, but we might be able to nudge things in the right direction now and again.

It must be frustrating to be these guys (the authors). In this book and others they have written, they show how fundamentally broken our approach to management has been ... and yet, if they were to take the next logical step and suggest, not adjustments to the system, but a new system altogether, their lifeless bodies would no doubt be thrown into the street outside Stanford Business School. But even if they are unable themselves to take that step, they at least have the satisfaction of laying the intellectual groundwork to help make it happen.

Here are the passages I've highlighted from the final chapter:

• Wells has shown that eyewitnesses make far fewer errors when presented one suspect at a time -- and there is no particular expense or technology barrier to showing people pictures or suspects sequentially rather than all at once. Yet although Wells has spent 20 years campaigning to convince police departments to use sequential lineups, only four out of over 19,000 legal jurisdictions have implemented this evidence-based practice.

Of course, many of the problems associated with adopting evidence-based approaches, or even just their findings, may come from the factors addressed in Switch: How to Change Things When Change is Hard.

• Peter Drucker: "Thinking is very hard work. And management fashions are a wonderful substitute for thinking."
• 1. Treat Your Organization as an Unfinished Prototype
• 2. No Brag, Just Facts
• 3. Master the Obvious and Mundane
• ... two economists walking down the street ... spot a $20 bill lying on the sidewalk. The first says, "Look, a $20 bill. Let's pick it up." The second replies, "It can't be a $20 bill -- if a $20 bill were lying on the sidewalk, someone would have picked it up by now."
• 4. See Yourself and Your Organization as Outsiders Do
• ... study of successful versus unsuccessful Silicon Valley start-ups found that in companies that survived and thrived, the CEO usually had a trusted counselor on the team -- while CEOs of unsuccessful firms usually did not. These counselors were typically 10 to 20 years older than the CEO, with broad industry experience, and were most valuable for helping CEOs recognize when they were traveling down the wrong path and a shift in strategic direction was needed.
• 5. Power, Prestige, and Performance Make You Stubborn, Stupid and Resistant to Valid Evidence
• ... wise leaders should begin by assuming -- no matter how flexible, smart, and open to facts they've been in the past -- that the mere act of stepping into a powerful position can transform them into stubborn, dumb, and evidence-resistant jerks.
• 6. Evidence-Based Management is Not Just for Senior Executives
• ... the chilling similarity between the Challenger and Columbia tragedies, where in both cases engineers had evidence that a problem was likely, yet people with more power but less knowledge ignored their warnings and sometimes bullied those who made them ... people who were "marginal and powerless" had "useful information or opinions" that they didn't express.
• ... under the Toyota system, assembly workers were actively involved in and responsible for tracking their own quality and productivity, coming up with ideas about how to improve them, and designing experiments to test their ideas -- all with remarkably light management oversight.
• 7. Like Everything Else, You Still Need to Sell It
• 8. If All Else Fails, Slow the Spread of Bad Practices
• ... a case can be made that when leaders are wrong -- and people don't have the power to reverse their commands -- that ignoring orders, delaying action, or implementing programs incompletely may be best for all involved.
• 9. The Best Diagnostic Question: What Happens When People Fail?

(A list of "best diagnostic questions" would be excellent. I would add What's Your Employee Turnover).

• ... forgive and remember... Organizations that forgive and forget keep making the same mistakes over and over again. Organizations that remember -- but blame, stigmatize, and punish losers -- create a climate of fear. Forgiving but remembering failure promotes learning without creating a climate of fear. Remembering also helps because when the same people keep making the same mistakes again and again (and others don't), it is a sign that these people need more training or are better suited for a different job.
• There isn't a shred of evidence that great leaders have magical powers that result in superior decisions and strategies, devised with little or no help from the mere mortals they tower over, inspire, and command.

What Is (Real) Leadership?

In the previous entry, we looked at mainstream management's obsession with leadership. Or rather, the myth that has been built up around the absolute power and omniscience of leaders as you move up the management hierarchy. The belief that leaders must be all-powerful god-kings is perfectly reflected in executive compensation: Starting in 1980, the average American CEO made 42 times as much as the average blue collar worker, then 85 times as much in 1990, and 531 times as much by 2000. I can't tell if the compensation dial was being turned up in an attempt to get better performance, or as a reward for the perceived improvements that were attributed to executives.

The belief that leadership is the primary factor and that we need bigger and bigger leadership is countered by evidence showing that "...their actions rarely explain more than 10 percent of the differences in performance between the best and the worst organizations and teams. Scores of more recent studies confirm that the link between leadership and performance is modest." Good leaders can have "modest" effects on the company, but truly bad leaders can have devastating effects -- the potential upside is mild, and the potential downside is great.

Hard Facts then began suggesting -- without going all the way and saying it -- that perhaps minimizing rather than maximizing leadership was the answer. This seems to agree with the main theme of Drive: people want to be productive. Yet most employees have had the experience of management that impedes rather than aids productivity.

Maybe what we need to do is the exact opposite of this insane cult of leadership: rather than making leadership as big as possible, make it as small as possible.

My most interesting leadership experience has been in running open-spaces conferences. Presumably, I'm the leader, and yet my job is just to explain how it works and then get out of the way and let it happen (I simply become another participant). Occasionally I will need to answer questions and even make decisions, but as more people become familiar with the system, they are equally capable of answering these questions or making these decisions (at the last Java Posse Roundup, we added "Ask Me" stickers to the name badges of anyone who had attended before). Mostly, I just convey information which anyone could convey. Last year, I broke my leg and was barely able to get around or show up for much of the conference, and it went on just fine without me. Being the "leader" of an open-spaces conference is just about the most minimal job you can find -- and because of that it allows me to do numerous other things (in particular, participate in and enjoy the conference) rather than being trapped in the "command center" as happens in traditional conferences.

So I ask: instead of continuing down this path of deep hierarchies and big leaders, what if we went the other direction: flatten the organization as much as possible -- making it completely flat if we can, and making the leadership positions as minimal as possible -- making them temporary or eliminating them if we can? I'll talk about how to minimize or eliminate leadership positions in the next post.

The Myth of Leadership

Chapter 8 of Hard Facts asks the question "Are great leaders in control of their companies ... and should they be?" and produces compelling challenges to the whole cult-of-leadership that has been the driving force behind much of the behavior of senior management in recent decades. They also make a strong case that  ridiculous executive salaries are not only impossible to justify, but they might even encourage bad behavior.

I will treat separately the question of leadership and possible alternatives. Here are my highlighted passages from the chapter (passages are bulleted):

• So much is written about leaders because we believe that our fate, and the fate of our organizations, is in their hands and ought to be. We talk and act as if leaders are all-powerful deities and devils who wield complete command over even the largest organizations and that the organizations are better off for this fact.
• In 1980, the average CEO in America made 42 times as much as the average blue collar worker, a ratio that rose to 85 times as much in 1990, and to a whopping 531 times as much by 2000. When Harvard Business School leadership researcher Rakesh Khurana asked corporate directors if CEOs are worth all that money, they reacted with anger and surprise, as if he had raised a taboo subject. Khurana found that directors held "virtually religious" convictions on the subject, which led them to dismiss any evidence showing that CEO quality is not a primary and potent cause of firm performance.
• Most people believe that leaders in more senior positions -- those higher up the chain of command -- not only have the right, but also have the responsibility, to make important decisions about and for those serving under them. People in higher positions are presumed to know what should be done and how to do it better than their underlings ...
• Making decisions and exercising power over others are also among the rewards that many people seek and enjoy by attaining higher-level positions.
• ... in many instances, organizations have failed because of excessive centralization and too much influence and control on the part of the leader.
• ... 60% to 75% of the employees in any organization -- no matter when or where the survey was completed and no matter what occupational group was involved -- report that the worst or most stressful aspect of their job is their immediate supervisor." These studies also show that "abusive and incompetent management create billions of dollars of lost productivity each year." Study after study demonstrates that bad leaders destroy the health, happiness, loyalty, and productivity of their subordinates.
• We have witnessed firsthand how poor leaders can drive skilled and motivated people out of their organizations and into the arms of competitors, or perhaps even worse, cause people to withhold discretionary effort even while remaining in their jobs.

Bad leadership can be devastating, but the best leadership will only occasionally justify outlandish compensation. Leaders should instead try to follow the hippocratic oath: "First, do no harm..."

• When Jeffrey Pfeffer published a review of research on leadership back in 1977, he found that although leaders do have some impact, their actions rarely explain more than 10 percent of the differences in performance between the best and the worst organizations and teams. Scores of more recent studies confirm that the link between leadership and performance is modest.

A friend and I have often joked that the best marketers are primarily talented at marketing themselves. It would seem that the most highly-paid executives are primarily talented at convincing everyone that they have the power to make a difference.

The book points out that these are half-truths. Having a bad leader really does make a big (negative) impact, and on extremely rare occasions you find an exceptional leader that really does make a big positive difference -- Steve Jobs is certainly an example, but is clearly a special statistical outlier.

• ... avoiding bad leaders may be a crucial goal, perhaps more important than getting great leaders.

If a leader is simply mediocre, people can work around him without too much pain and effort. But a bad leader can have a ruinously destructive impact.

• Members attribute team performance to their leaders even when the game is rigged so that leadership doesn't matter.
• Yet another reason leaders get too much credit or blame for what goes on boils down to money, power, and prestige. Senior executives, and just about everyone they come into contact with, have powerful incentives to act as if leaders are in control of organizations and ought to be. Leaders have enormous incentives for perpetuating the myth that they are in control, including the nearly $10 million per year that the average CEO of the largest 200 U.S. firms was paid in 2003.

The following is important, and I'm going to make a proposal (in a separate entry) based on this idea:

• One of the most consistent findings in the literature on decision making and performance is that the best groups perform better than the best individuals, because the groups are able to take advantage of the collective wisdom and insight of multiple individuals, while individual judgments reflect the narrower insights and skills of just one person.
• Leaders make mistakes -- all people do. But to the extent leaders exert tremendous control over their organizations, there are few or no checks or balances to reign in the errors. Most corporate disasters and financial scandals ... happened ... because such people were in leadership positions with so much control that no one could challenge them or raise questions. Placing so much control in the hands of a single individual violates the principle of checks and balances, a principle designed to ensure that any single individual, no matter how mistaken or how flawed, cannot do unlimited damage.
• ... the rest of the people have no investment in actually carrying through on actions and choices they had no part in making. So ceding too much control to leaders may make those leaders feel better about the choices that get made, but almost certainly will leave the organization confronting substantial challenges in getting things implemented.
• ... having both perceived and actual control over what happens in our lives is essential to the mental and physical health of all human beings, and it causes us to try harder, too.
• Because leaders succumb to the same self-enhancement tendencies as everyone else, magnified by the adulation they receive, they have a tendency to lose their behavioral inhibitions and behave in destructive ways.
• So sometimes the best leadership is no leadership at all.
• Leaders often have the most positive impact when they help build systems where the actions of a few powerful and magnificently skilled people matter least. Perhaps the best way to view leadership is as the task of architecting organizational systems, temas, and cultures -- as establishing the conditions and preconditions for others to succeed.

 The preceding group of passages all point in a clear direction, which you might be able to anticipate based on ideas in some of my previous posts.

• ... even as you are instilling that confidence in others and projecting that you know what you are doing, it is imperative to avoid succumbing to your own hype, believing your own press, and as a consequence, suffering the downsides of believing in your own omnipotence.
• Gruenfeld and her colleagues show that when people are put in positions of power, they start talking more, taking what they want for themselves, ignoring what other people say or want, ignoring how less-powerful people react to their behavior, acting rudely, and generally treating any situation or person as a means for satisfying their own needs -- and that being put in a position of power blinds them to the fact that they are acting like jerks.

In short, making someone a leader sends them a clear message that we are not all in this together. Indeed, how they react to that message is how you know truly what kind of leader they are.

• There are ways, of course, to partly avoid the problems of believing one's own hype, and they all entail reducing the differences in how people are treated -- for instance in pay and perquisites.

If that's true, it also suggests the opposite: the bigger the pay difference, the more likely you may be to believe your own hype.

• ... as news travels up hierarchical levels, each messenger changes it a bit to tell the boss a happier and happier story. ... the mum effect ...
• Leaders can also reduce power differences by selecting, training, and promoting senior managers who see themselves as serving rather than dominating others in the organization.

But such leaders, I think, might be too humble to effectively market themselves to the directors that are doing executive searches. The best leader for the company is hardest to find among the noise of the good self-promoters.

• If you look at the best evidence, however, there are more times when intruding on and evaluating underlings -- or just watching them work -- damages performance. There are also many other times when, even though leaders may believe they are making things better, they are simply useless.
• The best leaders know when and how to get out of the way.
• ... leadership entails getting things done through other people, and sometimes the best way to do that is to let those people use their knowledge and talents without interference.
• In practice, convincing leaders to get out of the way is difficult because their power and earnings are related to how many people they oversee, because so many people believe that a decision made by a leader is better than one made by a subordinate, and because so many leaders believe that they make better decisions than their subordinates and believe that if they watch, question, and nag their subordinates, those underling will perform better.

It is excruciating when the authors can get so close to coming to a conclusion -- they lay out all the dots, but can't seem to connect them for some reason -- possibly because they teach in a management school, and can't quite bring themselves to the conclusion that could challenge everything their school, profession and everyone who graduated from it stands for.

• ... leaders exercise the greatest personal influence over a company or group when it is young, small, or both.
• A second, related myth about leadership is that a skilled leader can manage well in any company or industry.
• Becoming an effective manager requires deep knowledge of your industry, organization, people and the work they do.
• Bringing in an outsider typically bolsters performance only when the company is in financial trouble or senior management are stealing money or lying to shareholders...

Change or Die?

This is the title of Chapter 7 of Hard Facts, wherein they question the cult of change and change for the sake of change. There are lots of management books that start with the assumption that everyone wants to be in a constant state of change. Making changes isn't exactly wrong, but it's a half-truth and the advocates of constant change hide the difficult side effects and unintended consequences. This is compounded by the problem that "change" is an easy thing to package and sell, so management consulting firms are incentivized to do so. They can be long gone by the time the change produces negative effects or even fails -- and failure happens more often than success.

Here are my highlighted passages from the chapter as bullet points (un-bulleted passages are my own comments):

• Hewlett-Packard (HP) insiders told us they were flabbergasted to discover that HP had done no research on how consumers viewed Compaq products until months after HP CEO Carly Fiorina announced that the two firms intended to merge. Fiorina was unhappy to hear that consumers viewed many Compaq products as among the worst in the marketplace and saw most HP products as far superior to Compaq's products. But this information was only considered long after HP's top management and board were committed to the merger and after Fiorina had told her top management team that she didn't want to hear any dissent over the merger plans.

A manager who says, "Here's what we're doing and I don't want to hear any discussion about it" is one that causes disasters when they make mistakes, because information already gets filtered through rose glasses as it goes up the management hierarchy. Another example of a bad merger happened long ago (in computer time) when Borland bought its competitor, Ashton Tate (the creator of DBASE), for 1 billion, an amount that, in short order, Borland wished it still had. The thing was, Borland already made the superior product and customers had been hemorrhaging from Ashton Tate to Borland anyway. There was no reason, and no advantage, to buying such a sinking ship. There were plenty of people (myself included) who could see the merger as a huge mistake as soon as we heard about it, but no one at the top wanted to hear the things that they really, really needed to hear. Borland today is only a wisp of what it was, and what it could have been. The knowledge was there, within the crowd that was the company, to make good decisions, but senior management "knew better." That's what I mean when I talk about the destabilizing effect of the hierarchy -- the hierarchy tends to discard good knowledge in favor of "desirable" knowledge.

• Although start-ups born with high-commitment HR practices fail less often and perform better than other start-ups, those that later try to switch to such "better" practices perform worse and die at higher rates compared with those that don't make such changes, apparently because of the disruption caused by trying to switch.

Some things must be grown into the culture from the beginning. The Goal and Fundamental Operating Principle are not things that are easily changed later, because everyone who joins the company is (usually unconsciously) buying into those values.

• Business process reengineering (BPR) was ballyhooed as a cure for every organizational problem in the late 1980s, but even BPR guru Michael Hammer -- who got rich off the movement -- now admits that most BPR efforts failed to achieve their goals and that the failure rate of such programs was almost 70 percent.
• ... there is no solid evidence that using layoffs rather than less draconian methods to cut costs increases performance, and there is plenty of evidence that involuntary reductions in force damage both displaced employees and "survivors."
• ... naysayers who oppose every new change and innovation will be right much of the time. But companies that never try anything new or introduce new products or explore new strategies are almost certain to die in the long run ...
• A study of 947 acquisitions between 1970 and 1989 shows that bigger firms that buy smaller firms (with stock) suffer substantially (25%) inferior returns over a five-year period compared with otherwise similar firms that do not attempt to grow through acquisition. Shareholders of acquired firms do benefit if they sell their stock quickly (especially in the first few months) ...

It's interesting that, elsewhere in the book, they give Cisco's formula for successful acquisitions, one point being that Cisco would only buy significantly smaller companies. However, Cisco put a lot of effort into getting everything right and is thus a statistical outlier.

• A survey of 232 IT executives by consulting firm Robbins-Gioia found that 51% viewed their ERP (enterprise resource planning) implementations as unsuccessful. A survey of 365 IT executives by the Standish Group showed the typical enterprise software project took about twice as long and cost about twice as much as originally planned, and over 30% were cancelled before completion. Botched implementations happened at Boeing, Dell, Hershey's, Nike, and Stanford. FoxMeyer claimed that a poor implementation drove them into bankruptcy.

That's only the tip of the iceberg. Many estimate the losses ensuing from the bad design of Java Enterprise Edition (J2EE, the foundation of many of these systems) was easily in the multiple billions.

• Studies of TQM implementations show that, especially after TQM became a fad, companies would often talk about their TQM programs, perhaps do some TQM training, but not take steps to improve the quality of their products or services.
• ... by the mid-1990s, it was clear that most BPR projects were falling far short of their goals and many projects entailed layoffs of people who were actually needed.
• A University of Colorado study of S&P 500 firms between 1982 and 2000 showed no link between downsizing and subsequent return on assets. A Bain study of S&P 500 companies found that the 158 firms that used layoffs primarily for cost cutting suffered modest drops in stock price.
• A Right Associates study found that 70% of managers who remained in downsized firms reported that layoffs were followed by lower employee morale and trust in management.
• Nokia started as a lumber company in 1865; sold tires, rubber boots, and computers along the way; and eventually became the leading producer of mobile phones. Hasbro started out selling carpet remnants in 1923 before becoming a toy company. Marriot started out selling root beer in 1927 before evolving into one of the largest hotel chains in the world.
• ... firms that introduced a steady flow of new innovations were more likely to live longer than those that didn't.

I argue that some firms can see the opportunities to change direction and introduce innovations, and others can't. The ones that can don't fundamentally change who they are when they change direction and/or innovate. Their core values are what allow those changes.

• Some changes may not be improvements because your organization is actually already doing them -- you just don't know it!
• We hear this all the time when we talk about, for instance, mergers or ERP implementations -- that the other companies weren't as clever, didn't do as much or as good planning, didn't have the same level of talent -- a lot of stories about why "this time, it will be different."
• One organizational change not evaluated here is whether there is a first-mover advantage for companies that introduce the first product in a category or are first to enter a new market -- the evidence is so inconsistent that developing simple guidelines for management action isn't possible yet, at least in our judgment.

When I wrote Thinking in Java, I received many bits of sage advice about how, since there were already 200+ books out there on Java, it was a wasted effort. I did not have the first mover advantage.

• ... economists and managers routinely use models that assume human beings have unlimited information-processing capacity.
• We've emphasized throughout this book that traveling through life with an attitude of wisdom -- the ability to act with knowledge while doubting what you know -- is the single most important quality that a leader, adviser, or team can have for practicing evidence-based management.
• ...IDEO's guiding philosophy was "enlightened trial and error outperforms the planning of flawless intellects."
• A change poses lower risk if there is an option to discontinue if it becomes clear that the new idea, however well intentioned, was a mistake.
• ... escalating commitment, the idea that once we make a choice, particularly if that choice is both public and consequential, we are reluctant to change.
• Decades of research have shown that escalation of commitment is a widespread phenomenon and one that is hard to avoid or overcome.
• A second process may be even more pernicious: the tendency to avoid looking at the consequences of past decisions.
• We have been almost completely unsuccessful in getting board members to go back and review those decisions, not so they can assign blame, but so the can actually learn something.

There's a tool for this, the retrospective, created by Norm Kerth and explained in his book Project Retrospectives: A Handbook for Team Reviews. Note the intelligence in using the term "retrospective," which doesn't immediately make you think of "blaming people for failure."

• One way to avoid feeling bad is to avoid confronting your mistakes ...
• ... if people never understand how and why the milk got spilled in the first place, the odds of recurrence are higher than if learning had actually happened. In the worst cases, people make the same mistakes again and again, but pretend as if the same old problems are brand-new challenges.
• When fear and insecurity are rampant, reviews of past actions and decisions often degenerate into a game of blame and punishment, which squelches learning and continuous improvement.
• Despite all the warnings about resistance to change, the belief that "organizational change is difficult and takes a long time" is a dangerous half-truth. Renowned organizational researchers including James March and Karl Weick show that organizations are surprisingly adaptive and can change quickly and easily.

Also see my review of Switch; those authors make the point that change isn't so hard when you want to do it; also that it's possible to understand the dynamics of change well enough to make it happen much more smoothly.

• Change happens when:

1. People are dissatisfied with the status quo.
2. The direction they need to go is clear (at least much of the time) and they stay focused on that direction.
3. There is confidence conveyed to others -- more accurately overconfidence -- that it will succeed (so long as it is punctuated by reflective self-doubt and updating as new information rolls in).
4. They accept that change is a messy process marked by episodes of confusion and anxiety that people must endure.

• Creating a self-fulfilling prophecy is one of the most powerful things you can do to increase the odds of success, regardless of the success rate experienced elsewhere. To avoid the dangers of operating on blind faith, however, confident actions need to be punctuated by episodes where people openly discuss their doubts and uncertainties and update to incorporate new facts into what the organization does.
• ... "the urgent drives out the important,"
• ... emotions are contagious, including the emotions of confidence and enthusiasm. So, belief in success helps motivate and inspire others, and creates a climate where effort, and thus success, is more likely.

A True Exit Strategy

The term exit strategy is about money -- investor money, specifically. If you put in a bunch of money (or effort) into a company, how do you get it out? It reduces the company to a single financial transaction, and emphasizes that your only goal in creating or investing is that moment when the big money changes hands.

There's also a fundamental assumption that businesses will only grow. All plans and contingencies are around growth.

When there's a pull-back in business, everyone panics, like they could never have expected this to happen (because they didn't). It doesn't matter that it's happened before, to lots of other companies -- we never thought it could happen to us! It's a reflection of how we, as a society, handle sickness and death (mostly with denial). We have no mechanism in place for pulling back and slowing down, only for charging forward.

It's even harder with a deep hierarchy, where only happy news reaches the top. People do kill the messenger, so the messengers filter the data, and the king only hears that things are going well -- at worst, pretty well. The truth only reaches the top when things are close to collapse, or collapsing. If the king doesn't like the news, he or she can declare that bad news is unacceptable (sound ridiculous? That's exactly what Carly Fiorina did when she began hearing -- long after the wheels of the HP-Compaq merger were turning -- reasons that it wasn't such a good idea).

Maybe we're hard wired for denial, and loss aversion. Perhaps it served us well in more primitive times, kept us from giving up when believing that things would get better might just give you enough inspiration to escape the clutches of that saber-tooth tiger. That strategy doesn't work so well in a world that is constantly changing around you, faster and faster.

What inspired this? Video stores. Netflix (and, let's face it, downloading) is killing them. And now there's Redbox, so any store with space enough for a refrigerator can also be a video store. The days of video stores are numbered, especially with the forced (via competition) reduction or removal of late fees.

Is there any way the local indie video store can evolve "downward" to move out of video-only and diversify to include more stable forms of income? Eventually, perhaps, replacing the video section with their own Redbox? (The general stores with a video section have a mental advantage here, because videos are already just one thing that they sell).

I have no idea how to think about this problem because, like everyone else, I've been immersed in a culture that can only imagine growth. Perhaps we should study the military -- they've always had strategies for falling back.

Reinventing Management

A reader pointed out the Management Innovation eXchange, a new site by McKinsey & Company with the goal of "Reinventing Management." Looks like it might have some valuable resources.

There's something in the air, clearly.

Why Do We Work Together?

In the previous post I talked about the two pieces required to describe a company: the Goal and the Fundamental Operating Principle. In a further epiphany, I realized that these concepts could be used to describe the motivation and behavior of all businesses, and to answer the fundamental question I've been asking myself: Why do businesses exist? At first, this sounds like a silly question, but what I'm looking for is more primary: Why do we work together?

To reinvent business, we must ask this question -- especially now that technology has given us so many new choices about how a business can be organized.

The answer tells me the primary reason for the existence of a company. And if we can align the needs of the company with the needs of the individual, we can minimize management (this is a good thing, especially if you're a manager, because your life gets a lot easier and more pleasant) and create a self-evolving, self-stabilizing company.

Why do we work together? Because:

Working together is better than working alone.

That might seem like an oversimplified declaration, but I mean it to describe the Goal of all companies (actually, for all group activities), and it also creates a test: If you find yourself thinking that working alone would be better than working for your company, then the company isn't achieving its goal (and perhaps you should work alone, or for a different company). And trying to force a company to exist when everyone is better off working alone won't make sense.

It also tells you that there should be some benefit to working together. In the industrial age, you simply needed more hands and backs to make something happen, but in the information age more brains is not always better. Sometimes independent contractors producing isolated but interconnectable components is the ideal solution (more than one person has put forward a business plan based on maximized independence of employees; essentially a company that's a group of contractors). But there are benefits to working together:

• Knowledge sharing
• Learning together
• Group problem-solving
• Helping each other get unstuck

Applying last century's industrial-style management techniques to these activities has not worked out.

So the Goal of every company is to make working together better than working alone. What about the Fundamental Operating Principle? This came from a flash of insight. The fundamental operating principle for all groups, and especially companies, is:

We succeed together, and we fail together.

This concept also gives an important reason why it's better to work together than work alone. It also shows why so many companies get out of whack, even to the point where they are at war with themselves.

If I work on my own, I succeed and fail on my own. Succeeding is great -- I get all the benefits of success -- but failing can be disastrous. However, if I join a company and we succeed, we all benefit (at least in theory). If the company heads downhill, there's a buffering factor; the company presumably stores resources in good times that can be used during difficult times, so I might still get a paycheck long after I would have stopped making money working on my own. We are significantly more loss-averse than we are adventurous, so this has great appeal.

A second essential aspect of succeeding and failing together is the attitude it creates: if you're failing at your job, then I'm failing at my job, and it benefits me to do whatever it takes to help you succeed.

When a company or the people in it work against this principle, all kinds of unpleasantness ensues. The biggest of these is I succeed, you fail. One of the worst impacts of this comes from disproportionate executive pay and benefits; the bitterest example is a CEO who gets a bonus for laying people off. Since we seem to be hardwired to seek fairness, there is a strong message sent when senior management gets heaped with benefits when the company succeeds, but if things go bad their risk is minimized (they might still get rewards!) while everyone else might lose their jobs. It's very difficult to believe much of what management says when you feel that kind of disparity.

If the message is "your failure is not my problem," then there is no value in working together; I don't get helped if I'm failing and I don't help you if you're having problems. In that case, I'd be better off working for myself.

If you can see and incorporate "we succeed together, and we fail together" as the core of your fundamental operating principle -- and make it real, not a trick to try to get people to work harder without you putting any skin in the game -- then your company can be a genuinely great place to work. Try to game the principle and you are in for the usual unhappiness we see so often in the business world.

Fundamental Operating Principle

Mission statements are commonly ludicrous and often the object of ridicule. No one remembers them. They feel like marketing lies. This is probably because the mission statement is typically created under duress, as part of what you need to do to convince people to give you money.

"Huh? What's a mission statement?"

"It ... uh ... briefly sums up everything about the company."

So they make one up, check the box, and it becomes official.

There are two primary things you need to express about your company. The first is the goal: what is your company going to accomplish? The second, commonly neglected, principle is how you're going to accomplish this goal. I'll call this the fundamental operating principle. This is the thing that people can repeat to themselves when they have choices about how to go about doing something.

Let's look at some examples.

Google. The goal seems to be, "Provide Services (to sell ads)." But the fundamental operating principle is "don't be evil," which people have often considered to be the mission statement. It didn't work as a standalone mission statement because it only says how, and not what. The benefit of breaking a mission statement into goal and fundamental operating principle is that you can quickly see if you missed one.

Netflix. The goal is something like, "deliver electronic entertainment." The fundamental operating principle might be, "Exceptional trustworthy employees."

Zappos. Goal: "Shoes through the internet." Fundamental operating principle: "Employees ask forgiveness, not permission."

Microsoft. Goal (as expressed by Bill Gates): "A computer on every desk." This is interesting for many reasons. For one thing, it seems to focus on desktop computers, not cell phones or tablet devices or music players (where Microsoft has consistently failed. The XBox is a special case that seems to have required exceptional effort and investment -- but could also be seen as a variation of a desktop computer). Maybe Gates got out because the "desktop computer" goal had been mostly accomplished and he didn't think he could change the company's direction. But it's also interesting because Microsoft doesn't sell computers, so the goal might have confused employees. What's their fundamental operating principle? It's hard to tell; I don't think it's explicit, probably because it's something like, "win at all costs," and that's not something you want to write down. But people in the company learn by example, so you have those that follow and end up fighting each other within the company in order to win, and those who don't want to fight and just go along. The result is what we've seen for the past 10 years: lots of confusion and muddle, and there may be winners and losers within the company but the customers are mostly not winning anything.

Burning Man. Their example made me realize that there is both a goal and a fundamental operating principle, because Burning Man makes them so clear. The goal is "Radical Self-Expression." The fundamental operating principle is "Leave No Trace." Notice how the fundamental operating principle governs choices about behavior, and can often produce an appropriate tension with the goal. Here, for example, it is permissible to create a sculpture that you plan to set on fire, but you must also make sure that the result doesn't leave anything behind. What's brilliant about "Leave No Trace" is that it can be interpreted both physically -- don't leave any trash behind -- but also emotionally/spiritually, as in, "don't do something that hurts someone (leaves a trace in them)."

Open Spaces Conferences. Goal: "Maximize Interaction." Fundamental operating principle: "Do what works best for you." The reason open spaces work so well and so effortlessly is that the fundamental operating principle relies on The Stability of Autonomy -- the best interest of the conference aligns with the best interest of the individual, so there's no energy needed to force the alignment, as is required by most companies. But if the interests of the individual and the interests of the company can be in natural alignment, the effort required to force that alignment (a.k.a. management) can be minimized or eliminated.

For what other companies can you describe goals and fundamental operating principles? Can you describe these traits for your own company?

Next up: a flash of insight produces the fundamental operating principle for all group activities -- and explains how companies can get so alienated from themselves, work at cross-purposes and get less and less efficient.

Stefan Sagmeister's Guide to Happiness

Within this TED talk, the designer Stefan Sagmeister gave his points for happiness, many of which apply to his working life:

• Complaining is silly. Either act or forget
• Thinking life will be better in the future is stupid. I have to live now.
• Being untruthful works against me.
• Helping other people helps me.
• Organizing a charity group is surprisingly easy.
• Everything I do always comes back to me.
• Drugs feel great in the beginning and become a drag later on.
• Over time I get used to everything and start taking it for granted.
• Money does not make me happy.
• Traveling alone is helpful for a new perspective on life.
• Assuming is stifling.
• Keeping a diary supports my personal development.
• Trying to look good limits my life.
• Worrying solves nothing.
• Material luxuries are best enjoyed in small doses.
• Having guts always works out for me.

He has a second more recent presentation making the case for taking time off to reinvent himself (which he does every seven years, and which I'm doing now).

The Stability of Autonomy

Suppose I create a business, and you want to join it. To do so, you must give up some portion of your autonomy. Autonomy is too weak a word, because it suggests a kind of positive choice. If you join my business, you relinquish much of your power over your own life -- certainly while you're at work, and in many situations when you are away from work (I can expect you to do work at home, or call you into the office for urgent tasks, etc. I might even claim to own any ideas that you come up with, regardless of when and where they happen).

Why would you give up this much power to me? Well, I created the company so I know what's best for it, and if you want to join, I'll pay you but I get to decide what everyone does. Not only do I have the vision, but I know what's right and best for every aspect of running that business. Or, if the company is big enough, my senior management team knows what's right and best. After all, I, or my team, spend all our time thinking about this stuff, steering the company in the correct direction. Basically, you give up your power to me because I know what's right, and you don't. So you have to do what I say.

If you don't agree with what I say, you will spend some of your paid time in resisting -- probably small resistances, and possibly unconscious. I once worked at a company where we were on salary, and were expected to "put in the necessary time," but we were also required to fill out a time card every week that said we worked 40 hours. I took this to heart and tried to work about 40 hours a week, so as not to defraud the time card. But I was also a little slow in turning in the card, which held up the bean counters a little bit, and gave me a small secret thrill because I was sticking it to the man.

This resistance is expected. Economists call it shirking, and that's what I expect from you. Basically, you're going to be lazy and unmotivated, and my job is to compel you or incentivize you to work. Everybody is prone to shirking and resisting. You're basically a bunch of truculent five-year-olds, and I have to herd you.

Naturally, this system doesn't scale easily. As we grow and add more five-year-olds, we must add more herders, who spend all their time struggling with the five-year-olds, who shirk and resist in turn. As the company grows, more and more of the effort is spent in these activities, and less and less can be devoted to creating and selling products. That's just the way of the world, and that's why you need to give up your power to me.

Now, if you were not five-year-olds -- if you were adults, and could take care of yourselves somehow -- then we could do things differently. The world is filled with management systems for five-year-olds, so I don't really know what this would look like. But I could hazard a few guesses. Let's take things to extremes and see what happens.

First, I'll assume that you don't need anyone to watch over you. You're an adult, so you can take care of yourself. Also, you know how to work together so that we're all rowing in the same direction. And you know how to act in the best interests of the company. Basically, in the "adult" version of a company, we can get rid of everyone whose job it is to watch over the behavior of everyone else. Which is nice, because that saves money. In fact, a lot of money.

What happens, you might wonder, if this "adult" company accidentally ends up with a five-year-old? If we don't have all those systems in place to manage that person, won't they run rampant and cause all kinds of expensive trouble? Yes, that might happen -- although I suspect it might happen far less if everyone's an adult. We'd probably be able to spot most of those before hiring. And if we did hire someone who turned out not to fit, then there are probably ways to encourage them to leave. For example, Zappos has a policy of paying new employees $2,000 to quit if they're unhappy with their jobs.

I don't know yet how decisions would be made, but it seems like we should be able to push a lot of the decisions "down to the appropriate level," so that the people who are affected by the decisions, make the decisions. More complex decisions should be made as close to their effects as possible. Perhaps there are gaming techniques that could speed these kinds of decisions if they need to be done in a group. If everyone has a strong vested interest in maximizing the success of the company, there must be some way to crowdsource the decision-making process.

If we continue using the old system, it means that (A) you give your power to me, which makes you unhappy and unproductive and (B) I become isolated from what's really going on in the company, and start making more and more disconnected decisions. My control of the company seems, for this reason, to tend towards instability. If, on the other hand, we are constantly getting feedback from people who are closest to the problems -- and not just information feedback, but control feedback -- then the company will react more and more quickly to both problems and opportunities, and tend towards not only greater stability, but greater efficiency and productivity.

What Questions Would You Ask?

(Originally published 12-26-06; supplemented by reader comments)

The questions you forget to ask when you are interviewing for a job, but wish you'd asked after taking the job. The first seven are mine, the rest were added by readers; the reader contributions vary in tone significantly. I've done some editing, and removed the ones that are software-specific.

• If I want to buy something like a book or a tool, how does the process work (how hard is it?). What's the cost limit before the approval must go up the management chain?
• What's the noise level like during the day?
• How many meetings am I expected to attend, and how long do they usually last?
• Is there a dress code?
• Can I work from home sometimes?
• Does it matter when I work, as long as I come to meetings?
• How many projects have succeeded/failed in the last five years? To what do you attribute the failures?
• What is the complete job hierarchy, how many people are at each level, and exactly which level is my position?
• Does management adequately communicate the business side of things to employees?
• Does management shut down push back from employees, or do they listen? (and vice-versa).
• How does management deal with deadlines? Missed deadlines?
• Are things are changeable? (That's the difference between a bad state of things being an excuse for itself and an opportunity for growth).
• How serious is the company about knowledge sharing and prevention of one person from owning critical information (also known as the guy who can never take his holidays)?
• How do you determine who gets to work on which projects and how they are rewarded for their work?
• Would you rather hire someone from outside the company for a new position or promote someone who already works for you?
• How do you help people keep their skills up to date and develop new ones?
• How does an employee who has an idea for a new product or project present it to someone with the authority to move it forward? If it's a good idea, what sort of reward or recognition is typical?
• What percentage of management came from within the nuts-and-bolts areas of the company? How long ago were they doing that?
• How are you required to report weekly status?
• How many times do you have to report about your time? (Like in three different systems: once for upper management, once for HR and once for the team lead, plus a team meeting just to repeat verbally what has been written).
• Please describe in detail your level of bureaucracy. What bureaucratic systems are in place? How heavy is reliance on rote procedures vs. active understanding?
• Are mistakes encouraged or punished?
• Is responsibility commensurate with power given, or do employees have no power and all the responsibility?
• How often is upper management criticized from below? How often have they changed the way they conduct business in response? Does upper management actively and vigorously solicit such criticism?
• Why are you in business? Fundamentally why? To make money? To perfect the state of the art? To improve human condition? To be the top dog? The answer to this will dictate the atmosphere in the company and the day-to-day feeling of the place.
• Are people here considered allies or resources?
• Is anything being done to improve the flexibility of job schedule? For example, if someone wants to work 3 days a week for some part of the year, is that negotiable? Is the company open or shut to such possibilities?
• What is done to protect the health and sanity of employees?
• If the company is private, are there any designs to go public and "cash out"? Is there an intention to sell the company?
• Are you a transparent organization? How can I see this?
• How important is consensus? How can I see it?
• What is viewed as good work and acceptable work in this culture?
• How do you try to increase employee productivity?
• Do you use wikis for communication and documentation development?
• The Cluetrain Manifesto is a good resource.
• Why are you here?
• How easily does change spread in this organization? From the top, or from the bottom?
• What are you hoping I won't ask?
• If you could change one thing about your company what would it be?
• What is the personnel turnover? Both the long term turnover, and for the past 6/12 months.
• Who is my immediate superior?
• Am I supposed to work with another department without any prior notice by my superior?
• Who is my immediate subordinate (if there's any)?
• Can another department 'borrow' my subordinate without informing me first?
• Do you guys go to lunch together? (If the answer to that isn't positive then I know I don't want to work there).
• How are the paychecks funded? (Product sales, advertising, venture capital -- if something happens to the economy, will you be out of work?)
• How much debt does the company have? (How much of your time is paying off the company's past mistakes?)
• What is the company's annual revenues? How many employees? (Do a mental calculation of revenue / employee, if it's less than $100,000 / employee, then that's a potential warning sign.)
• What is the company's net profit/loss?
• What is your top technical risk or problem right now?
• What is your top social or organisational risk or problem?
• What is the turnover rate for people coming in to similar positions at similar seniority levels?
• What do you like about this place?
• What DON'T you like?
• Some companies want employees to sign away all rights to the ideas they generate, whether at home or on the job. What's your position, assuming no conflict of interest?
• Writing books, articles, etc. Are there any issues if I continue to write the occasional book or article?
• Any issues around blogging? Obviously, it's important to exercise good judgment when discussing work-related topics.
• Moonlighting -- any policy on occasional part-time work for others? (Assuming no conflict of interest or effect upon my work at my day job).
• Training/Education/Books/Tools. Do you pay (within reason) for employees to attend conferences or training? What about books and software tools?
• Do you care how I work as long as I finish my work on time?
• Can I take a holiday if I finish up my work early?