A Perfect Mess: The Hidden Benefits of Disorder | Eric Abrahamson | Talks at Google

Thank you, everybody. Thanks for having me. One thing Jimmy didn't say is I grew up in France, but rest assured, I too hate the French. The other thing that made it that I grew up in France was my dad was an artist painter, and I decided not to go that way. So my only way of sublimating my artistic needs is through PowerPoint. So you'll excuse any PowerPoint geekness that I engage in. So the name of the book is A Perfect Mess. And people typically want to know initially why I would write a book on messiness. And I'm a professor of organizations at Columbia University. But I'm messy and that creates a certain tension in my brain because I'm supposed to talk about all about organization. And if you saw my office, it looks like a bomb went off in it. It's really, really, really bad. So you can check out my website if you don't believe me. But basically, I got a bit, as a joke, interested in this subject, and then I wrote a 60 page tome on the subject of messiness. I don't recommend it unless you need something that's less dangerous than Ambien or something like that. But I wrote this article and then I ran into, very much by chance, a journalist who was interested in randomness. And together we wrote this book, A Perfect Mess. So that's a bit the genesis of the book. I'll talk, but you should feel free to jump in. Challenge me, tell me I'm an idiot, do whatever you want, Agree, disagree violently, the only rule will be no violence. So the book's been, to my great surprise, very, very, very successful. Not that I care very much. It's not false modesty. I like working on things, but I don't very much care what happens to them afterwards. But it sold 70,000 copies, it was 100 top 100 in Amazon and Barnes and Nobles, it got reviewed, got on TV and the foreign rights were sold in 22 countries. So that's the most astonishing part of it, that messiness is so interesting internationally, I think, because capitalism generates lots of stuff, lots of objects, lots of ideas, lots of possibilities, and that to a certain extent, we all struggle with messiness to a certain extent. Oops, oops, oops, oops. So the agenda for me is going to. I'm going to talk about the cost of order, because typically people think more order is better and they don't think about how costly it is to achieve order. Then I'll talk about efficient mess in space, efficient mess over time. Here I'd like you to think A little bit of the career of Arnold Schwarzenegger, who went from being a weightlifter and could have gone somewhere else, judging by this photo, to being a movie producer to being. Or movie actor rather to. To being governor. And he's notorious because he does not have. Until he was governor. He does not have a very clearly defined agenda for his days. If something's interesting, he'll spend two hours on it. If someone's a bore, he'll spend two minutes on it. So order and space are people who have very. In time rather are people who have very, very rigid and clear agenda. Whereas mess and space are people who are much more that improvise a lot more in what they do. I'll talk about mess and creativity. The two are highly, highly related. We'll talk also very briefly about mess and power. Talk about how mess is a source of power and mess and beauty and different types of messes and how beautiful they can be. So if I can get through all of that in 45 minutes. If we can't, we can't. And we'll leave time for questions. And you can ask in the questions. All right. The argument in the book is not just about messiness in desks. It's messiness in organizations. It's even messiness in industrial districts, entire industrial districts. It's messiness in entire governments. So some of the arguments about messiness may be somewhat scale invariant. That is some of the characteristics at work at one level of analysis may work at multiple levels of analysis, Typically, and this is somewhat amazing. And people think very much about the benefits of order, but they greatly underestimate the cost of order and the cost of reorganization. There are overly organized organizations we call. Often call them bureaucracies. A tremendous amount of money spent on organization, 50 billion on consultants. But we know that a lot of firms fail during large reorganization. 70% of major changes are unsuccessful. So order is both costly and risky in terms of changing order. Have a number of examples of organizations that were growing very, very rapidly. Not very, very organized by standards that we would accept in terms of bureaucratic standards. They're then taken over. They're rendered much, much more orderly. And then they fail. And one of the reasons they did so well is because they bore so little cost of organization that allowed them to be so profitable. And once all the HR departments and standard operating procedures and processes are put in place, all that cost of organizations makes these organizations less efficient. Six Sigma is another one of these current fads. It's unclear whether the cost of order to achieve six Sigma is greater than its returns and we can talk about that if you want. So the takeaway here is to do a cost benefit analysis of order to think about. All right? Are we somehow over investing in order? And if we're over investing in order, that means we should be less orderly. In other words, we should be messier. This, for instance, is IBM's organizational chart. Now do you know what a matrix organization is? IBM has a seven dimensional matrix and in talking to a lot of people who work at IBM, they can't find anybody. It's just so complicated, it's so complex as an organizational form that it's almost impossible to find anybody. So that might be an instance of an over organized organization and we could debate about that, but that's been my experience in thinking about IBM. So high level mathematics, the organization returns minus the organizational cost me is equal to the organizational benefits and oops. And if the organizational benefits are less than zero, then we should have more mess T plus one. So here's an example of mess. Mess in space. This is a, a very interesting woman who runs. A very prestigious magazine. I can't tell you who she is. She would absolutely kill me if she knew that I was showing this photo to you. This is not exactly what I have in mind. The book is not advocating being a complete pig and being completely, optimally, completely messy. The book argues more that there's some kind of mid level between complete order and complete messiness. And that mid level, that sweet spot, is where systems work the most efficiently. You'll note there's an ibuprofen bottle here which I think is not completely unrelated to the state of our office. So the book, so the claim of the book is not going to be simply be messy, be messier, it's going to be more. How do you find that balance of order and messiness that is optimal for the situation that you're looking at? We of course love Einstein. Einstein was a very messy individual. He's the one who said if a cluttered desk is a sign of a cluttered mind of what then is an empty desk? Extremely, extremely messy individual throughout his life. And a little bit of our poster boy for messiness to indicate that people can be very, very messy or moderately messy and be quite efficient. This is a funny finding that we. So we did this web survey, not very scientific, but it's the best that we could do at this time. And we find in our survey that people with an orderly desk reported spending 60% more time finding things than people with a disorderly Disk and a bit. I mean, I think it's worth sort of thinking about why that might be the case. But frequently people create very, very complex files and have trouble finding what they're looking for in their files. There's also certain returns to scale to messiness. If you let 10 things form on your desk, you only have to do one trip to file the 10 things. If you clean up your desk every time something hits the desk, you have to put it away, take it out, and so on. So there are certain returns to scale in terms of messiness. And then there's the more common sort of experience where you say to yourself, I'm going to put my keys here so I don't forget where they are. And then you forget where you put your keys. Right? An example of being orderly in some way, but then forgetting where you organized yourself. But anyhow, our survey doesn't seem to indicate that more orderliness in desks leads to finding more things more easily. So there's this returns to scale of order, people who can't find what they've lost in their order. There's another sort of argument is the opportunity cost of a mess. I don't know if you have an urgent task and your desk is messy, what are you going to do first, the urgent task or clean up your desk? Probably the urgent task. So there are a lot of efficiency sort of benefits to having a messy desk. And what's necessary is to do some kind of cost benefit analysis. What is the ratio of time it takes you to clean up your desk to the time you save by having a clean desk? And typically people who have extremely clean desks will claim that the cost benefits analysis is in their favor, but there's not a lot of evidence to support that. I did a mouse is I did a computer simulation of a moderate messy. So this is the degree of organization, this is the returns to organization. So how much time you put gets you more organized. And this is a speed of task to completion. So this is when you get deeper into the pocket, it takes you less time to organize. And did little like nerdy three dimension simulation. And you see under most parameter settings there's somewhere in the middle where we find that a moderate level of messiness is optimal. And this is simply a simulation which says when you work it creates a mess, mess slows you down and you can either stop working and clean up or, or continue working. And that simulation seems to suggest a sweet spot of moderate messiness across different parameter settings. The GUI computer interface is an interesting instance. It was invented or discovered or designed by actually a gentleman who studied desks. And he studied 10 desks, five messy desks, and five very, very clean desks. And he found that, in fact, people made much greater use of messy desks. Messy desks have everything you need right in front of you. They're easy to grab, easy to find, easy to combine. And he used that metaphor in order to create these GUI interfaces. I don't know if I can do this. There's one. How do I start this video? Well, anyhow, this is a new type of GUI interface which is completely based on the notion of messiness. They have piles. It allows you to look at what's in piles, it allows you to put a document a little bit at an angle in a pile so you can see what you're looking for, you can move piles, you can recombine piles, and so on. So they're a kind of interface that completely takes advantage of messiness as a tool, as an advantage in managing your computer. Oh, it's here. You want to watch a little bit of it? No, not going to happen. All right. But if you get a chance, it's kind of fun to go to bumtop.com and see how they've used messiness in terms of an organizing principle in the definition in the GUI interfaces for desks mess over time. So we talked a little bit about Arnold Schwarzenegger. A person who's very orderly over time is a person who has every 15 minutes of their day mapped out. Schwarzenegger is a person who never had that kind of a calendar. And the argument is that if you can have that kind of flexibility in your calendar, that if you can improvise because your calendar is not so fixed, you can take advantage of many, many more opportunities. That if you had an extremely orderly calendar, the same holds for organizations. Now, there are a lot of studies about the strategic planning in organizations. Most of the studies seem to indicate that strategic planning is good for organizations. But surprisingly, most of those studies are carried out by strategic planners. So there's a slight bias there. Those studies that are not carried out by strategic planners look something like this. Business performance, extent of strategic planning. If you can fit a regression line through that, good for you. But there's really, from what we can tell, no pattern in terms of the extent of, of strategic planning and business performance. And once again, I think the argument is if you think of the Soviet Union and if you think of five year plans, you think of sort of the inflexibility of strategic plans and how difficult it is for companies to adapt to new opportunities that come along. And you think of organizations that have much looser plans or no plans at all and that are capable of adapting very, very quickly to customers. It's a bit the same thing as Arnold Schwarzenegger, but now it's at an organizational level of analysis. Why doesn't strategic planning not work? Most of strategic planning is done without a real grasp of the facts. Less than 50% of CEOs could recall last year's sales for the firm. The only thing in studies that managers agree upon is whether they're in a big or a small firm. So a lot of strategic planning is is done with very, very bad data. And that's maybe one of the explanations. Long term plans act like blinders to new opportunities. And to a certain extent that is going to hurt firms. And their argument that long term sustainable competitive advantage is not sustainable is the third argument here. I use Google frequently as an example. I go to more and more more on the Google website and I give an example of all the different things that Google is trying out. Almost like an evolutionary process where you have a lot of variation. Some of it will work, some of it will fail. Some the stuff that works gets selected and retained. So the new theories of strategies say that most competitive advantage erodes. Initially you don't do as well because you're starting out, then you start doing a lot better and then you reach some kind of point where you've exploited all the new competitive advantage. And the new theories of strategies say you should then jump on a new learning curve that would take you up. The issue of course is how do I get on the next learning curve? Let me skip this. And basically the argument is that you have to do a lot of small experiments. Now those experiments, should they be highly correlated? No. All right, you want highly uncorrelated experiments. So that sort of a mess of experiments really so that one of them will come up and have a chance of succeeding in bringing the company forward. So I use Google maybe rightly or wrongly, you'll have to tell me as a company that experiments a tremendous amount. And I judge this from the website in the hope that they're going to be able to hit a couple of these blue lines that will take them on the next learning curve. The post it. So if investing in a mess of innovation, you're looking at the cost, not the rate of failure. So you can have a very high rate of failure. What you want to do is you want to kill the failures very, very, very quickly and Let those parts, those experiments that seem to work, take off and become the next learning curve. Mess and creativity. One Nobel laureate called it the principle of limited sloppiness in experiments. And the argument there is to be sloppy enough so unexpected things can happen, but so sloppy that we can't find out what was the cause. So a lot of great inventions or a lot of great discoveries, rather, have occurred because an experimenter had a very, very controlled experimental condition, but not so controlled that something didn't slip in. There was a little bit of mess that slipped in. And when that little mess slipped in, that's when. So you have Louis Pasteur. I can get these slides to you. I won't go into great detail, but you have Louis Pasteur. You have Alexander Fleming, who had a very messy lab and let some stuff get into a petri dish, and he noticed that bacteria had invaded. And from that he deduced the existence of penicillin. There's big academic debate whether this is, in fact, how it happened or not. But it makes a nice story. But it's this principle of limited sloppiness. And there's instances, instances after instances of kind of sloppiness. And you can think also of an organization as one that holds too much of a certain type of employee and needs to let in certain kinds of sloppy employees or uncontrolled employees or different employees, and could create sort of innovative solutions more generally, messengers, creativity, because it juxtaposes things that would have otherwise been separated by order. That's something that we find at the level of desks, that we find at the level of organization, that we find at the level of industrial districts. That order tends to separate things. And when it separates things, we fail to see combinations that could be very. Now, I use. I'm from New York, so I use the example of the pizza bagel. All right? It's a bagel with pizza on it. And it emerged because you had a Jewish community next to an Italian community. So you have a mess of communities in New York. And only because you have that mess can you have that juxtaposition of two culinary traditions and come up with the pizza bagel, right? But this fun example. There are a lot of. There's a couple of example of Nobel discoveries. One of them is by this famous Nobel laureate. His approach is he'd have a desk, it would get very messy, and then he had a big roll of paper, and he would roll out another desk on top. And then he'd start another mess. And then when that got too big, he'd roll out another. And he'd do this for a month until he got sort of head to the level of his head. And then he'd clean up everything at once. Returns to scale, to cleaning up mess. And it's in one of these exercises that he found two papers. One posing a problem, another one offering a solution without a problem. And he saw the two come together as an idea that could be pathbreaking. Sent the articles to both of the scientists that got in touch. And in the story this creates another Nobel laureate winning experiment. Or I don't know how Google does it, but in a lot of organizations, people who serve the same functions, stay on the same floors, work on the same floors. And when you do that. So if you have marketing, finance, operations, sales and so on by floor, you tend to have an order of people like this. But certain organizations mix up randomly and messily the kinds of people they have on every floor. And that allows for a lot of creativity because someone from sales can bump into someone from finance, someone from marketing can bump into someone from sales. And again, something that would have been separated by order is recombined in an interesting kind of fashion. All right, so the Nobel discovery companies that mix functions by floor, we find that industrial districts with more varied companies tend to have more innovation. Again, the argument is that when the companies are so different, they tend to find new kinds of combinations of products and, and be much more creative than industrial districts that have very much the same kinds of companies within it. So this is something that may be scale invariant in terms of creativity. That what MES does is it juxtaposes things that would have been otherwise separated by order and therefore it allows for new kinds of recombination and new kinds of invention. Does that make sense? So that's the part for creativity. I've talked a little bit about efficiency, a little bit about creativity. One more thing, why this book was almost not written when we wanted to write the book, what do you think the reaction was? Where will this book fit into a bookstore? So this is absolutely insane. You can't write a book because it's not going to fit into an existing system of organization. Is it a science book? Is it a management book? Is it a self help book? We couldn't answer that question. So they said, well, you can't write the book now in truth, innovation or, or art, according to Heidegger, creates new categories. It doesn't fit into existing categories. So when you come up with something new and you have a system of categories, it won't fit into the existing categories. And therefore you get much less innovation unless you've got a flexible system. Mess and power. We all know the technique. If you create a massive mess in your work life, then it's very hard to fire you because only you know where everything is. If you're writing code, you have an advantage to writing messy code and not documenting it very well, because then it's harder for you to be replaced. So this is an example of mess and power with respect to governments. A system dominated by power is autocratic. One that's without power is anarchistic, really. Democracy is messy enough to let people do what they want, but not so messy that people one person's freedom will get into the way of another person's freedom. So again, democracy in this way of thinking is a moderate level of messiness. And then moderately messy. Business firms have the same advantage. If you create a mess, you're the only person who knows where things are. They can't fire you. Good for you, bad for the firm. Another example of a very powerful mess is messy organization Al Qaeda. This is a network map of the people who were involved in 9 11. And it's very, very hard to destroy a messy organization like this. It's very easy to destroy a very well organized organization. You kill the top, you kill the second level. But a messy organization is very, very hard to take out, particularly if it's being attacked by a very orderly organization. It's not clear what nodes you need to take out of this system to be able to destroy it. It therefore makes it a very powerful system in terms of its capacity. Finally, mess is beautiful. A Jackson Pollock painting. Again, it's a question of taste. And there are ugly messes, but they're also incredibly beautiful messes. This is a Frank Gehry building. Again, question of taste. But I think it's the messiness of the architecture, the lack of organization in it that makes it such a beautiful building. I think I'll put this. There was a question of whether God was messy. This is a woman who found a random pattern on her toast of the Virgin Mary and sold it on ebay for a lot of money. So again, who knows? That's all, folks. So what I've said, in short, really, is that the book doesn't advocate complete messiness. It advocates moderate levels of messiness. If you're completely orderly, you spend all your time creating order, no time to work. If you're completely disorderly, you have no capacity to work. So there's something in between a moderate level of Messiness, that's going to be the most efficient. I gave some of the arguments for why order is bad. Order is costly. Order is risky to change. I also gave some of the benefits of messiness returns to scale. You let things pile up and then you put everything back at once. You don't have to interrupt yourself 10 times to refile things. You can let a mess form and then put everything back at the same time. So certain benefits having to do with messiness. We talked a little bit about messiness and creativity, how messiness allows us to juxtapose things that would have otherwise been separated by order and that by being put together in the context of the mess. The mess allows for new creative recombinations and innovations. Talked a little bit about mess and power, how if you create a mess, you're the person who controls the mess and it's difficult therefore to oppose you. And finally talked a little bit about mess and beauty. Compare Soviet architecture again, with its very, very similar buildings repeated against and again. And the kind of ugliness that that creates. Or think of a very, very neat apartment that looks like nobody lives there. And think about moderately messy art, moderately messy architecture, moderately messy homes that allow the expression of the people there to come out. All right, so that's sort of my summary and my quick summary, my cliff notes of the book and shoot. I have, you know, please ask questions and I'd be glad to talk about them. Yeah.

My background is in entertainment. I steward in mainstream network television. And that fun little advantage graph is what a pilot season is ultimately.

Yes, making a bunch of little innovations.

But network television is probably over orchestra.

Yes.

The studio, the network production company, all sort of converging on the writers to sort of over organize it.

Like how.

Where is the sort of sweet spot in terms of something. I mean, network television is very big.

Yeah.

It's possible to have a sort of bright level of messiness because we certainly dealt with box when I was there.

Yeah. Finding the sweet spot is, I think, is very contingent on the situation. This is where I'm going to dodge the argument a little bit. It's very contingent on the situation. I think it's something that you arrive at trial through trial and error. You try and be a little bit too organized, you see the problems. You try and be a little bit too messy, you see the problems. So by trial and error, you find a moderate level of messiness that's not too costly, that allows for creativity, that allows for some of the goodies that we talked about. But in part the reason to write the book, to read the book rather, is to look at a lot of different situations where moderate messiness seems to work very well. And maybe from that glean some of the characteristics of moderate messiness that are the most efficient. But that's my story and I'm sticking to it. Yeah.

Did you look into at all why people fear messiness?

Yes. I mean, it's very interesting. We found that 60% of people feel extremely guilty about their messiness. And so I had to write the chapter on the history of messiness. And messiness in our Judeo Christian tradition is seen as bad. It's actually satanic. It's what undermines the religious order. So from the very, very beginning, it has very, very bad connotations. Then under the Industrial revolution, you get the rise of engineers. And engineers believe that orderly systems, orderly machines are going to be more efficient than. So then they take the ideas of machines and they apply them to social systems. So they say, if a machine works well, why not think of employees as cogs and of managers as engineers? And why don't we run an organization like a machine? And then they don't stop there. Also, they then go on to applying the SIM principles to the family. We talk a lot about the gilbreths who had 12 kids and who ran their family like a machine. They had them take baths and at the same time listen to German lessons and so on. They had all kinds of. They used Robert's Rules of Order whenever they had a conversation at the table and so on. So this conception of order is good, gets brought into the organization, gets brought into the family even, and gets considered as very, very important historically. And I think that's the root of why we feel such doubt, such insecurity, such guilt around messiness, when in fact there's not a whole lot of evidence that it's inefficient, that it's bad and so on. But that's, I think, the historical roots. Any other. Any other questions? Yeah.

Last weekend I was taking this time management workshop. How do you make a cleanup of your inbox 0 every few days yourself? You have like to do lists and project plan and all that. And it's sort of against what you're talking about.

Messy.

So is this sort of a contradiction?

Well, again, the whole time management literature doesn't, usually doesn't think about the cost involved in setting up your time management system. So of course, if it's free to set up your time management system, to run it, to change it when your schedule changes, and so on. Then it might be efficient, but it's costly to set up a time management system. And most people who set up a time management system don't stick to it and ultimately go back to right where they were. There's also another example are professional organizers. Professional organizers will come to your house and they'll charge you between 5 and $6,000 to organize your house. And the question is, are you going to get five or six thousand dollars of returns on that investment? Particularly since most people who are organized become disorganized over time. So very frequently it's so stupidly simple. People think, God, it's good to be organized. It's good to have time management or professional organizers or bring a consultant in to organize or organization and so on. But they don't think about the cost of doing that, the risk of doing that, and the benefits of doing that. And once you start weighing the cost versus the benefits, in a lot of instances you see that time management is really a loss of time, not a benefit of time, because they're not counting the time for their creating their time and maintaining their time management system. Six Sigma. Do you do any Six Sigma here? Thank God. It's another system which is extremely costly, not only to put in place, but also to really, really implement. And it's not clear that the cost of implementing that system yields benefits for that system. So again, many instances of people who want to organize you, but who ignore the cost of organizing you. Yeah.

I'm wondering what you think about the potential psychological costs of not being organized in terms of. And their effect on productivity. And I'm thinking, you know, whether it's sort of innate or learned or whatever. A lot of people do feel, they just feel more able to think clearly, et cetera, when things are not messy. Okay, you talk about that or do you see that as a potential benefit of cleaning is not just your ability to find things on your desk, but sort of a peace of mind.

A peace of mind, A kind of Zen state where everything is, well, lower anxiety. Again, you have to ask why is there the anxiety in the first place? Is it for historical reasons or for efficiency reasons? It's probably for historical reasons, but grooming.

And that sort of thing are really common sort of display of health and that sort of thing.

Yes.

Organizes like. I wonder if some of it is just innate, unnatural.

Well, there's actually interesting in psychology, the big five dimensions of psychology, and two of them are predictive of messiness. People who are more open to new experiences tend to be more messy and People who tend to follow rules less tend to be more messy. The flip side of it is people who are close to new experiences and, and people who tend to follow rules tend to be more orderly. Now people who have those psychological characteristics will tend to feel uncomfortable around messiness. People who don't have those psychological characteristics will feel perfectly at ease. I feel at ease in a mess. I in fact sometimes catch myself messing things up because I find them too orderly. So I think there's a psychological difference in how people react to mess.

So would you say from that people should strive to develop those characteristics that make themselves feel more comfortable around mess or simply choose the level of messiness that feels more appropriate to them?

I think both. I think both. I mean, I think certainly we find companies that have purging days or companies that punish employees that are messy. There's one famous example of this woman who got fired because she had three personal objects on their desks instead of two. Right. If you're passing rules like that, you're selecting against people who tend to think expansively and to not follow rules. Right. So I think you have to be. I think you have to be very careful how you apply these kinds of rules in a lot of these organizations. Also we found that these get clean days, people open a drawer, they take everything that's in and they shove it into a drawer and they close it and then you have a clean desk. And that's how they. So I think. So there's the psychological aspect and then there's the technical aspect. The technical research shows that people who have non routine jobs tend to be messier than people who have very routine jobs. So if you're constantly being bombarded by new kinds of tasks, you're going to tend to have piles of different things and so on. If you work at the motor vehicles department, for instance, where it's always the same task again and again, you can be very, very orderly. So I think there's some technical aspect that sets the level and some level of comfort psychologically that sets the level that's right for you in terms of messiness. But there's a lot of research to be done. My history is I've done the research first, usually, and then some journalist has ripped it off. So I thought I would rip myself off first and then I would do the research. So that's a little bit why I'm following this order. But it's a great question. Any other questions that come to mind? I've convinced you. Moderate messiness, mess is best.

Good.