Organization Posts

In the second part of my interview with Eric Ries, we discuss (1) acquiring customers without launching and (2) opening up board meetings to the entire company.

At IMVU, Eric and the management opened up board meetings to the entire company. Why?

  1. To give people the information they need to do their jobs.
  2. To teach everyone in the company to think like the CEO.
  3. To prevent employees from gossiping about board meetings.

And more!

I’ve synchronized the audio with some simple slides below. That’s my favorite way to consume the audio. You can also find a transcript and stand-alone audio below. Please let me know if you find the transcript useful.

Read on to learn what kind of employee Eric used to “show the door” at IMVU…

Slides: Opening up board meetings (pdf)




Audio: Opening up board meetings (mp3)

How do people find out about our product if we haven’t launched?

Nivi: And this gets to a second topic, which is you guys weren’t doing any of this really in public, because you had not launched a product, right?

Eric: Amen.

Nivi: Nobody knew who you were, but people were, at the same time, were using the product, so how did you do that?

Eric: I got a question today which was something like, “I’d love to follow your advice about not having a public launch, but we need to get early beta users for our product launch. How can we do that if we are not willing to talk to bloggers? Nobody really knows who we are.”

I think a lot of people have that attitude, that without PR, you just can’t get any early customers. Again, we have got to start with, “What is the goal of early customers? Why do you want them?”

If you are charging from day one, one of the reasons you want them is you actually want to make money. You want to show that your business is viable. But even if you are not charging money, you have a need to find out whether your business is viable, whether you have that minimum viable product, whether the business model, at the end of the day, is going to work.

For that, you do need customers, and you do need to be putting customers through a product experience that will give you that information. But you don’t need a lot of customers. I think that is where people get confused.

For a big fancy launch you can get hundreds of thousands of customers to show up for one day. But for metrics analysis, generally a cohort of 100 people or 1,000 people are plenty to learn from.

If you change your goal from, “How do we get the maximum number of customers,” to, “How do we get the minimally sufficient number of customers to learn what we need to learn,” new possibilities get opened up to you.

Acquiring customers on $5 a day

Eric: For example, at IMVU, we practice the $5 a day AdWords campaign. I was the VP of marketing in those days. If I actually knew anything about marketing, I would have known not to try this. By traditional marketing standards it is considered crazy to spend only $5 a day, but we had a pretty low budget and we really were pretty frugal.

I discovered that in those days you could buy clicks for five cents a click. But to me, $5 a day meant every single day 100 human beings are coming to try my product.

If you think about that from a beta testing point of view, especially if you look back at the old days of software shipped by CD, getting 100 people to try your product is actually a lot and you can learn a lot from that. And at 100 people a day, you are in good shape, just at that tiny, tiny level.

The risks of doing that are really quite low. I think a lot of engineers have this idea that once you put your product out there in public, the investigative journalists are going to find out about it and write about it and we are going to lose control of the story. Let me tell you. You should be so lucky.

IMVU was a top 1,000 website in the whole world before it got any press whatsoever. We were making millions of dollars a year. The press was writing about newly funded, venture backed competitors that had no traction whatsoever; because those were the guys sending them press releases.

It was frustrating, and psychologically you want to have that cover story on WIRED that you can send home to mom, but you know what? We did not start this company to have good vanity covers printed about you in the press. We were there to serve customers and serve them well.

Running experiments under a different brand name

Nivi: How do I run experiments, if I already accidentally got that TechCrunch article and I…?

Eric: Yeah, I am sorry. You are not doomed, but you are going to have to go waste energy later cleaning up the positioning that you put in that article, which is undoubtedly wrong.

Nivi: Right. There is that aspect of it, but do you think you should, just basically pretend that article never existed, or do you run tests under a different brand name?

Eric: That is not a bad idea. Especially on the iPhone, I see this because of Apple’s stringent release process where there is this huge delay before you can actually bring things to market.

And also because people want to get into the top 25. That is where all the action is in the Apps store. There is a lot of competition to make sure that on the day you launch your app you get all the right coverage lined up and all the stuff happening.

People feel like they don’t want to do a bad launch under their real brand name, because that will harm their ability to do the proper launch later and get to the top 25.

But there is no law that says you can only bring out products under one brand name. I strongly, recommend that to people if you are very concerned about your precious brand. I think most startups are way too concerned about the power of their brand. They should be so lucky to get some kind of brand going.

Even still, bring it out under a terrible name. I specifically recommend people bringing products out under brand names that they hate so that they won’t ever be tempted to make that into their real official brand name and then become afraid to experiment with it.

You have got to be bringing products out under a brand that you feel comfortable experimenting in. Then once you find the right formula, there are two possibilities.

Either you will be able to port that product over to your new brand name and it will be great, or the product concept you brought up under that bad brand name will be so powerful, you just can’t get people away from it.

It is too sticky and you are stuck with it. But congratulations! You are successful! Is it really so bad that you personally don’t like the brand if customers do like it? I think it is not so bad.

Running pricing experiments in public

Nivi: A friend of ours has a popular subscription based product that they don’t charge for and now they want to start charging for parts of it for the premium model, and they want to find the optimal pricing strategy. How can they run those experiments in public and in secret? What would you suggest to them?

Eric: I would actually not be afraid to run them in public. It is hard for people who are afraid of what the worst possible thing that could happen is, to do this. But I think it is good to just try it and get over it.

What happens is, it is true that customers don’t generally like the idea that one customer got charged one price for something that somebody else got charged a different price for the same item. So there is some risk when you do different pricing offers in a split test.

But in my experience, there are two mitigating factors that make it not so bad. The first is it is actually incredibly difficult for most customers to figure out that is happening, especially if you only do it in a limited time window.

For example, I am going to tell you a story that may not seem related, but bear with me. When we were at there.com, the virtual world company, we would do a lot of QA. That was a heavy QA company.

For hours every day we had QA testers sitting in a lab together running the virtual world software and testing to make sure that it worked. I remember one day getting called in to see about…

There was one tester. They were around a physical corner from each other. So you couldn’t see each other, but they were not more than 20 feet away. They were both engaged in this activity.

The guy called me down and he said, “I am in this dune buggy riding around with somebody in the virtual world and we are seeing this glitch. We are not seeing the same thing. Something is not right.” They were calling back and forth, trying to pin down what it was.

I remember sitting there really confused about what the problem was, because it looked like the two of them were sitting there in the dune buggy and everything was fine.

I walk around the corner to the other guy. I talk to him about what the problem is. I look at his screen. On his screen, he and the other guy are engaged in a paintball match. They are not in a dune buggy at all.

He was almost a mile physically distant in the virtual world from where the other guy was, yet their conversation was perfectly consistent to them and it never occurred to anybody to ask, “Wait. What planet are you on in this time that we are comparing notes?”

They had no clue that this was happening. I think, we totally tend to underestimate just how powerful the pull of what you see is to most people. They basically can’t imagine the world, any other way than the way that it is.

Entrepreneurs don’t have that problem, so a lot of times they don’t grasp what is true for customers. It is actually very unusual for the customers to go onto a forum and post, “Here are all the offers that I am being offered and exactly what I see. Does anybody else see the same thing?” Our natural assumption is that everybody else sees the same thing.

So you are not totally likely to get caught. That is a mitigating factor. It is actually not as bad as you think when you do get caught, because don’t forget; you have the power of the apology, especially as a startup.

If you screw up… You are going to screw up all the time. If you are a customer of a startup, your general experience is, “These guys are constantly screwing up.” What customers care more about than whether you screw up or not, is how you treat them when you do screw up.

They care that you listen to them and take them seriously more than if you always get it right. If they want to work with a company that always gets it right, they will go work with some premium giant company that really has a very carefully constructed customer experience.

If you get caught doing this thing, you can always say, “We are so sorry. We were experimenting with this pricing. We didn’t mean for this to treat anyone unfairly. And if anyone was treated unfairly, we have gone back in the records and we are going to give them all double the money back for the thing that happened,” or whatever you have to do to make it right.

That is OK! It is really not that bad. What happens then is people say, “Wow. These guys are serious about making sure that we get treated fairly.” Meanwhile, you get to keep experimenting.

Opening board meetings to the entire company

Nivi: Yeah. You have talked a couple times on your blog about how you opened the board meetings up to the entire company and the positive benefits of that, and people’s perceptions of negative benefits.

Eric: Yeah. Well this is not something that a lot of companies adopt. This is considered pretty crazy.

I don’t know if it’s that most people are actually afraid of giving the whole company information they need to do their job, because it might lead them to judge the top management harshly, but people judge you harshly whether you give them the information or not, from my point of view.

Just give them the information! Your pathetic attempts to hide what’s happening don’t fool anybody.

Having been on both sides of that divide, I can tell you I never felt like I was being successfully fooled. And if you do manage to fool me for a limited time, I’m awfully pissed. My point of view is: you want people to have the maximum information possible.

You need to do it in a trust-building way. You’ve got to make sure the people you’re giving it to understand what they can and can’t do with that information, and they understand that they need to keep company secrets confidential.

If you don’t trust your employees to keep company secrets confidential, you’ve got bigger problems and you should go address those problems first.

There is some board business that has to be done in secret for legal reasons, so it’s not true that absolutely every meeting that any time ever happened at the board level is open to the public.

Employees have critical things to say in board meetings

Eric: But the interesting part about board meetings is the strategy conversation where you present progress, show data, and you make discussions about what should happen next. And that’s the part of the meeting I strongly recommend people open up to their employees.

What we did is we actually had a board of advisors and then a board of directors that was a subset of those advisors. We would convene the full set, advisors and board, at nine o’clock in the morning and we would have a maybe two-hour strategy conversation followed by maybe a half-hour or one-hour private board meeting.

For the strategy conversation the rule was: every employee can attend. We did this up until we were a 50- or 60-person company. We actually, physically crammed everybody into one room, and we had the employees sit around in as much seating as we could fit and the board members would sit at the big table.

It wasn’t a free-for-all, most of the employees were encouraged to listen, not to speak. But every once in a while the rule was that if someone had something they really needed to say, they could be recognized by the CEO and say their piece.

It was amazing. We would, occasionally have a board meeting where we would have a moment, where there would be data we were presenting to the board, and it would indicate that on a certain day, a certain metric went up and that was due to us launching that feature that day, or whatever our interpretation of what that data meant.

And not an insignificant number of times we would have an employee raise their hand and say, “Excuse me, but do you also realize that something else happened on that day?” Yada, yada, yada.

And occasionally, I’d be the one presenting! On the one hand, I’m really embarrassed. So I’m like, “No, I didn’t realize that.” This is a critical thing about running my own business I didn’t know.

But once I got over my personal embarrassment, what you would find is the board loved it! They’re like, “Thank God that guy was sitting in the room and could enlighten us about that. That changes our interpretation of what this means.”

And quite a few times I think we saved ourselves months of work by coming to a realization of something way earlier than we would have, because the right guy happened to be sitting in the room.

And yeah, occasionally you had an employee who’d make an off-color comment or say something that really shouldn’t have been said in front of the board, but people learn from those experiences. Most of the time most people had really substantive conversations.

Nivi: Did you ever get in a situation where some of the employees were like, “I don’t even care about these board meetings. I don’t even want to go?”

Eric: Yes, yes! We eventually had people who on occasion would beg me not to have to go to the meeting. And we eventually made them voluntary. For a while I was really rigorous, I said, “No, everyone has to be there. If I have to be at the meeting, you have to be at the meeting. Why do you think I’m any more privileged or unprivileged than you?”

Yeah, because board meetings are actually pretty boring. But when people are outside the room looking in — and you know most conference rooms have some form of glass — people can see what’s going on. They’ll come up with an excuse to walk by, kind of peek in. They will make up whatever crazy conspiracy theories are consistent with the data if they’re not there.

An, in my point of view, that’s such a source of waste: people gossip and there’s rumors and people don’t know. Let them be in the room, let them see how boring and mundane most board meetings are. So that for the occasional one where something actually interesting gets decided, let them be there to hear it themselves.

Everybody in the company has the ability to understand what everybody else in the company has to understand

Eric: There are some costs, definitely some down sides to doing it. One which took us by surprise was that, people can occasionally get confused about who’s in charge, we did occasionally have people — some board member would say, “You guys should really build feature X.”

Board members occasionally would just spout off about what’s randomly on their mind, and occasionally you’d have an employee get confused that that means the company is now going to go do feature X because board member so-and-so said so.

And that was actually good practice for us, to be a constant reminder that no matter who you are, no matter what it says on your business card, nobody gets to decide randomly that the company’s going to do feature X. Right? I don’t care if you’re the CEO or the lowliest person, we’re going to have a reasoned and considered process for deciding what to do.

Nivi: And it’s a learning opportunity.

Eric: It’s always a learning opportunity. The other thing that was hard for me personally was it’s hard to have your people who work for you see you be criticized in public. That was not fun.

Nivi: Hard for whom?

Eric: Well, it’s hard for me. My emotional reaction was like, “Wait a minute! I’m doing the best I can and now you’ve got to watch me get smacked around because I screwed something up.” But once I got over my personal emotional response to it, it was wonderful.

Because it humanized me to the people who worked for me — they got to see, “Oh, I see the pressure that he’s under” — but more importantly, when I needed something from somebody for the purposes of presenting to the board I could go to them and say, “Do you remember what happened the last time I didn’t have the right answers to these questions or I had shoddy this? You’re really going to send me in there with this? Come on, you’ve got to help me out!”

So it made us collaborators in creating solutions for the board rather than I’m constantly asking them for stuff and they don’t know why.

And I think, get over your own infallibility. We all make mistakes and it’s better for people to see what the real stuff is.

Nivi: I think you wrote about this on your site, basically the assumption is that everybody in the company has the ability to understand what everybody else in the company has to understand.

Eric: That’s right.

Nivi: The assumption is I have the ability to understand what the CEO has to understand.

Eric: That’s right. And that makes people uncomfortable, because sometimes we would say, “You have the obligation to understand what the CEO’s going through right now, because it’s going to impact the way you do your job.”

Some people would say, “I just want to sit in my narrow corner, do my little thing, and I don’t want to worry about what the company strategy is.” And we would show those people the door. We were really serious about that.

You really needed to have people who were… they didn’t have to be good at it! We weren’t asking them to be good at the CEO’s job, but we are asking them to understand why is the CEO making the decisions that he’s making. Because they’re going to have to make CEO-level decisions sometimes.

Sometimes the actions that have the biggest impact on the company’s performance are taken by people at the line employee level. They may not realize it’s going to have that big impact, but they are going to make those decisions. By the time the CEO finds out about it, sometimes it’s way too late to do anything about it.

We sure hope that the guy at the line level understands what the company strategy is and how his decisions impact, at least the best that he can.

Nivi: Yeah, I think maybe their decisions impact the company more than the CEO in the sense that if the CEO doesn’t come in to work, who cares? The company proceeds, but if the team doesn’t come in to work nothing happens.

Eric: OK, let me tell you: when the community manager takes a day off, you can have serious, serious meltdowns in the community if it happens to be the wrong day. That can have major impacts on the company.

Should we share bad news with employees?

Eric: I’ll say one more thing because this is a real effect that people are afraid of, which is that if you give people information about how a company’s doing, it can impact morale negatively. Sometime the company’s not doing well.

It makes some people have this idea that part of your job as a manager is to shield people from bad news or shield them from chaos. Because it’s not fair to them to have them have to do their job and also be confused about how the company’s doing. I just think that’s a really paternalistic attitude that we just need to let go as an industry.

If you want people to believe you when you tell them the good news, you have to sometimes tell them bad news. Otherwise, you have no credibility. And when there’s bad news to be shared, yes, it negatively impacts morale. But for a good reason, because things aren’t going well and we now need to rally the company around the fact that we need to change what we’re doing.

And there’s nothing like actually seeing the board say, “You guys have a major crisis on your hands that you have not yet understood,” to get everyone in the company saying, “We’re alarmed. We need to do something about it.”

That can cause some chaos, and that can be disruptive, but if you build trust and rapport with your employees then what you could do is you can sit everybody down for an analysis meeting after the board, which we would always do, and say, “OK, let’s talk about what we heard and what does it mean for the company,” and let people share their perspectives.

Let people say stuff like, “This says to me we need to cancel all our projects and completely retool.”

You need to get that idea out in the open because when somebody thinks that, you don’t want them to just unilaterally go execute on that plan! You want the opportunity to tell them and everybody else who didn’t have the courage to say the same thing: “No, we’re not retooling, but we are going to make some adjustments and here’s how we think about it, here’s what we’re going to do about it and here’s what’s going to happen.” That was pretty powerful.

Nivi: Thank you!

Eric: You’re welcome.

Nivi: I think that was great.

From The Human Equation:

“Seven dimensions that seem to characterize most if not all of the systems that produce profits through people.

  1. Employment security.
  2. Selective hiring of new personnel [especially screening for attributes that cannot be taught such as attitude and cultural fit.]
  3. Self-managed teams and decentralization of decision making as the basic principles of organizational design.
  4. Comparatively high compensation contingent on organizational performance.
  5. Extensive training.
  6. Reduced status distinctions and barriers, including dress, language, office arrangements, and wage differences across levels.
  7. Extensive sharing of financial and performance information throughout the organization.”

The author is Jeffrey Pfeffer, a professor at Stanford. He also co-wrote Hidden Value, which we covered in We don’t pay you to work here.

Implementing the whole system

Extreme Programming works well when you implement all of its practices. Most of the practices by themselves have too many flaws to be very effective. Each practice by itself may even have more disadvantages than advantages.

But all of the practices together work well. Why? For each practice, there are other practices that obviate its flaws. Wheels by themselves just roll a bit and fall over. But when you connect them to a car, the entire system can get you from Boston to San Francisco.

The practices in The Human Equation also work well when you implement them all. You can’t offer extensive training if you plan to lay people off when times are tough. That’s just a good way to waste money on training. You can’t offer employment security if you don’t hire new employees very selectively and if you don’t terminate the ones that aren’t effective.

If you’re searching for a magic incentive system to get high performance from your team, there isn’t one. If you’re willing to do the hard work of implementing a set of simple organizational practices, The Human Equation and Hidden Value have some suggestions.

“A raise is only a raise for thirty days; after that, it’s just your salary.”

– David Russo, VP of Human Resources at SAS Institute

This is one of my favorite quotes from the book Hidden Value. It explains why money by itself doesn’t motivate high performance. Money by itself can only motivate the quest for more money. A raise is only a raise for thirty days; after that, it’s just your salary.

We are motivated to perform when our work expresses who we are, when the business’ goals are intrinsically meaningful to us, and we feel that we are valued as people, not simply as economic agents.

But, even in startups, financial incentives and HR practices often treat us like economic agents:

“Consider the implicit values conveyed in the modern management practices adopted by many companies. Most firms today emphasize, among other things, the employee’s responsibility for being career resilient, employment at will and no-fault dismissal, pay for performance, downsizing to cut costs, and maximizing shareholder value above all else. What is the message any sentient employee takes from these practices? Pursue what is best for you, not the firm or the customer, adopt a free-agent mentality, and do not invest any more in the firm than it is willing to invest in you. The underlying values are crystal clear, even if they are never expressed in a formal way. In this sense, arguments by managers that value statements are irrelevant or inappropriate miss the point: All organizations have values; the only question is how explicit they are about them.

“And what happens when employees behave in accordance with these values? First, a rational employee is not likely to exert much effort in activities beyond what he or she is explicitly rewarded for. A ‘show me the money’ mood prevails. Second, a smart employee will be constantly alert for new and better job opportunities in other organizations—loyalty is for fools. Third, unless cooperation is explicitly monitored and rewarded, teamwork is viewed as optional… To resolve some of these problems, management’s job is to design ever more sophisticated control and incentive systems to ensure that the necessary teamwork occurs and that the loss of intellectual capital is minimized.”

The problem isn’t that money is a weak motivator. The problem is that money is a terribly strong motivator. By itself, money motivates the wrong people to do the wrong things in the quest for more money.

This is why Zappos pays employees to leave. This is why Tandem Computers didn’t tell employees their salaries until after they started working. In other words: we don’t pay you to work here—we pay you so you can work here.

Organizing around values, not value

The authors, Charles A. O’Reilly III and Jeffrey Pfeffer, both from Stanford’s Graduate School of Business, studied how eight companies, from Men’s Wearhouse to Cisco, ignore the pernicious assumption that compensation should be the foundation for management systems:

“First, each of these companies has a clear, well-articulated set of values that are widely shared and act as the foundation for the management practices that… provide a basis for the company’s competitive success. [e.g. Southwest's "Work should be fun… it can be play… enjoy it."]

“Second, each of these organizations has a remarkable degree of alignment and consistency in the people-centered practices that express its core values. [e.g. Southwest: "We hire happy people."]

“Finally, the senior managers in these firms, not just the founders or the CEO, are leaders whose primary role is to ensure that the values are maintained and constantly made real to all of the people who work in the organization… The senior managers in each of these companies see their roles not as managing the day-to-day business or even as making decisions about grand strategy but as setting and reinforcing the vision, values, and culture of the organization. Dennis Bakke at AES [a $2B company] claims that he made only two decisions in 1998, one of which was not to write a book on the company.”

Extraordinary results with ordinary people

The book’s subtitle is “How great companies achieve extraordinary results with ordinary people.”

Every rational company in the world is trying to hire the best people in the world. And all but one of them will fail at this task. There can only be one company with the best people. Hiring the best is a failing strategy.

Organizations must be designed to thrive with ordinary people. If businesses can thrive with the capabilities of ordinary people, they can also thrive with extraordinary people. Practices like Extreme Programming, that were designed for programmers with ordinary skills, work even better with extraordinary programmers.

Read Hidden Value for specific recruiting, training, information-sharing, and rewards practices that aim to exploit the capabilities of ordinary and extraordinary people alike.

“If people come for money, they will leave for money.”

James Treybig, CEO of Tandem Computers

Ed: This is a guest post by Kevin Meyer, the President of Factory Strategies Group, which operates Superfactory. He also writes an excellent blog called Evolving Excellence. In this post, Kevin describes how American Apparel unwittingly applies lean practices like short cycle times (concept-to-product in 8 days), integrated QA, cross-functional teams, and more.

American Apparel has long been one of my favorite companies. I love companies that think outside the box, ignore popular wisdom, and, in doing so, teach us some lessons.

Brief background on why I like them so much: this is a $500 million manufacturer of t-shirts, underwear, and the like. Typically low margin products, the kind of thing that usually comes from Asian and Central American sweatshops.

But not at American Apparel. This company makes over 1 million articles of clothing per week, from their one factory in Los Angeles and they grew 40% this year. They pay their 5,000-person workforce significantly above minimum wage (average is $12-$15 per hour), give them full subsidized benefits (such as high quality health care insurance for $8 per week), and they turn a profit.

This should embarrass the heck out of any executive who thinks he has to outsource in order to find effective labor. Or at least call into question his fundamental competence as a leader. If American Apparel can manufacture low margin clothing efficiently enough to beat the sweatshops (in California no less), then anyone should be able to. If they try hard enough.

Apparently some of the brass at the company have kept tabs on my blog as, about a month ago, I got a phone call inviting me to come down for a visit. It was everything I expected, and more, and in many aspects it rivaled the various Japanese factories I visited recently.

First Impressions

One of those aspects was the first impression. When we arrived at Saishunkan in Japan, we were greeted by a gardener who turned out to be the chairwoman of the $270 million company. At another factory we walked in to see the president of a $100 million company on his knees scrubbing the floor.

When you arrive at American Apparel, you see several massive multi-story warehouse buildings. At the business address is an open entrance with an old table and a visitor sign-in sheet. That’s it. It’s on the lower left of building in the photo below. No, not the far left… that would be the company store. The open gap entrance just to the right of the store… which looks like any of the other roll-up dock doors.

No fancy lobby with glitzy lighting and display cases, no plush waiting rooms. An open entrance with a guard and a sign-in sheet. For a $500 million company with over 5,000 employees.

As you can tell from this photo and the others, the buildings aren’t in particularly great shape either. But they serve their purpose, and do it well. They may need a coat of paint, but they’re clean, neat, and filled with a lot of happy brains.

The people

The value at American Apparel isn’t created by the building, or the machines, but by the people. So many companies—probably the vast majority—think of their people as a cost. Very few companies recognize the value of the brain that sits slightly north of their people’s hands. American Apparel realizes that the value of that brain more than offsets the traditional cost difference between their employee’s hands and a pair of hands in a sweatshop.

What value has been added by the brains of their employees? How about this: their highest volume product, a “deep V” t-shirt, was the idea of one of their shop floor folks. Or this: American Apparel makes and sells a variety of unusual products, such as dog sweaters and baby bibs. Why? Because their employees figured out how to design those products with the little remaining scrap that exists after cutting out the patterns for the mainline products. There’s still some scrap left, so they had another idea: create a machine that would combine and weave it into bikini straps and cords for hoodies… like in the photo on the right. There’s still a tiny bit of scrap material left, so that is sent to a recycler, who turns it back into yarn and thread, which is then turned back into cloth for more products.

In fact, sustainability is a big deal to American Apparel. How many of you haven’t embarked on sustainability programs because of their cost? Well, American Apparel recycles just about everything, obtains 30% of its power from solar cells on the roof (and they are looking into getting much more), and many of their trucks run on biodiesel. They buy as much organic cotton as they can… domestic organic cotton—they believe that the carbon footprint created by sourcing organic cotton from overseas is too much of a negative offset. Anyone interested in growing more organic cotton in the U.S., here’s your customer. So once again, if you think sustainability is too expensive, then you should be embarrassed.

Workcenters

But let’s get back to the people. Production takes place on each floor of all of these buildings by multitudes of 4 to 8 person cells (they call them “workcenters”). A kit of cut cloth is wheeled to each cell and they crank through it. A chart of metrics is maintained at each cell. See the column on the right, which I know you can’t read? It’s dollars… and reflects the dollar value of what the cell has created, and most importantly the actual dollar portion that they get to share. A form of piecework on top of a nice base hourly rate. Each cell has a quality control person, and other quality people roam between the cells.

Do you see that they’re smiling? It’s no wonder; they are valued and treated very well. Not only do they have a doctor on site, they have a full modern clinic. Back when there was a transit strike in LA, the company bought a couple thousand bicycles and created a bike loan program, with free maintenance, that still exists. How many companies have purposely eliminated phones on their manufacturing floor to cut costs? American Apparel has phones all over the place, and provides free calls—even long distance. That’s a major benefit to their primarily immigrant workforce.

And perhaps most importantly, the company actively solicits their employee’s ideas and recommendations, and they actively implement them. It’s no wonder their retention rate is over 98%.

Vertical integration

A few more concepts.

How about the importance of gemba? Administration, marketing, design, and other offices are scattered throughout the buildings. They are not adjacent to each other. So everyone must walk through the factory floor multiple times a day. Everyone is continually aware that they are in a manufacturing facility.

Another concept: complete vertical integration. Everything is done at this cluster of buildings, with the exception of some dyeing that is done a few miles away. Design is done, often tested by Dov Charney himself, and sent to the factory floor. Time from raw concept to a finished product in over 200 stores worldwide? Eight days. Compare that to the weeks and months it can take to send a container across the ocean. Did the Olson Twins wear something unique yesterday? A new design can be created and placed into stores almost immediately to capitalize on the brief craze.

Complete advertising development, through photographing the models, to final printing, is also done at the factory. The props for all of the stores are created and sent from this factory. When a new store is opened, the fixtures and initial inventory are sent, often pre-hung on hangars so the new retail clerks can focus on selling.

Since everything is created in one factory, they can react fast, and therefore the stores don’t have to maintain as much inventory. Shipments, globally, are smaller and more frequent. They have a unique way to balance raw material inventory: if raw inventory gets too large, they simply create a new design that will consume that inventory and sell it. A luxury many of us wish we had.

Are you embarrassed yet?

I’ve previously written about Sun Hydraulics, a 1,000 employee, $170M company, with no job titles except “plant manager”… the guy in charge of watering the plants in the factory. American Apparel is very similar. Sure there is the CEO, there’s a marketing department, and there are cell leaders. But not much more. I asked a couple of my hosts what they did and I got answers like “some strategy stuff, but then I also figure out how to hire people for the stores.” Basically whatever needs to be done.

American Apparel may be very altruistic, but they still realize they’re a business. A business that has to make money to continue to provide the stable solid jobs for their valued workforce. My hosts told me about some other companies with similar values, who focused too heavily on the “communal good” and soon went out of business.

And yes, I saw Dov. He’s been in the news quite a bit recently, thanks to a workplace atmosphere that would make some cringe. At the risk of offending the more sensitive among us, part of me applauds his guts to run the company as he sees fit, traditional rules be damned. It even earned him a great spoof on Saturday Night Live. By any measure, he is a retail genius and he has the smarts to see outside of the “must outsource to make clothes” traditional mindset.

But here’s my final and perhaps most important lesson: do what works. It’s that simple. Tools, even lean tools, are just tools. Leadership requires people. At American Apparel there are no cheesy signs with “Teamwork” and “Challenge” on them. There are no glitzy glass lobbies. There is no sign of lean manufacturing in the traditional sense, and they don’t profess to be lean. No heavy lean training of employees, no overwhelming visual controls besides the metrics charts at the cells, no Shingo Prizes or Baldrige Awards.

But there are bunch of people recognized and compensated for their knowledge, creativity, ideas, and experience. A group of people who realize that speed creates value, knowledge creates ideas, ideas create profit. They figure out what works, then they do it exceptionally well.

A 5,000 person, $500 million low margin clothing company, operating from a single factory in the least business-friendly state of one of the highest “cost” manufacturing countries. Beating the overseas sweatshops and still growing rapidly.

Are you embarrassed yet?

We don’t know a lot about organizing organizations.

We apply techniques like command and control, specialization, batch processing, hierarchy, and other other lessons we’ve learned from TV, our parents, previous jobs, school, the military, wherever.

But there are other ways to organize organizations. And we’re learning more about them every day.

Evolving Excellence writes about Sun Hydraulics, a company that’s organized in an extraordinary way:

“Their culture is really something to see.  A $170 million public company that manufactures high end hydraulic manifolds and valves, profitable since it was started in 1970, [with] six plants around the world employing roughly a thousand people.

“What’s unusual about that?  How about this:

  • There is no organization chart
  • There are no job titles or job descriptions
  • No performance criteria
  • No bonuses and no perks
  • No regularly scheduled meetings
  • No approval levels for capital or expense spending
  • No goals
  • No offices or high-walled cubicles
  • If the peers accept the idea, then “management” is presumed to accept it — hence the need for very little management
  • Every employee is simply expected to figure out where they fit

“There is one honorary job title: Plant Manager.  But it’s not what you think.  This facility, what amounts to a very large machine shop filled with heavy 5-axis CNC’s, has hundreds of live plants hanging from the ceiling.  The Plant Manager is the person in charge of maintaining the plants.”

Read the rest of the post on Evolving Excellence for some hints on how they organize under these conditions. More extraordinary organizations coming up.