This is probably one of the most important talks you will ever see if you are an entrepreneur in any discipline. It was the opening talk at 2011’s Business of Software.
Professor Clayton Christensen talks models of disruptive innovation, what the job of your product is and explains why there will always be room for disruptive innovation and entrepreneurs in the world. Trust us, you just don’t want to miss this talk – a classic.
Video & Transcript below
Learn how great companies are run
At BoS we run events and publish content that is highly valued by anyone trying to build, run, and scale a great software company.
Sign up for a regular dose of actionable and useful content:
Professor Clayton Christensen: Well thank you Mark. I apologize for having a tie. [Laughter] My wife actually made me sign a note that I would always do that just so that I don’t, offend her. You guys look pretty civil actually. I got to go, a number of years ago, to a meeting of the free software foundation [Laughter]. And I don’t know if you remember the first Star Wars movie when Luke Skywalker came in and met Hans Solo and remember that band that played there. There was like one of every conceivable type of humanity in the band and that’s what that conference would look like. But like I said you guys look more civil.
Just a couple of other things about me, being an academic is my second career. The first was I founded and ran a company with several MIT professors to make products out of a set of advanced materials that they had developed. The company has become quite successful but I always wanted to be a teacher. So as the company hit its stride and I was hitting age 40 with the support of a great wife decided if I was ever gonna do it then I had better do it. And so I bailed out with the No. 2 person in charge and I became a doctoral student all over again at the age 40 with 5 kids. So, it was really stupid. [Laughter].
But in some ways I’m really glad that I did it in that sequence, that’s because I brought with me a set of puzzles that I probably wouldn’t have thought about had I just been a beginning-to-end academic. The other thing that you should know is that about a year ago I had a stroke and the quite came from somewhere enlarged itself right inside there where you fabricate your speech and your ability to write and it just killed my ability to speak. So over the last year I went out to Logan airport and got a copy of Rosetta Stone for English and have been working through that. So you will see that I’m speaking more slowly than normal. And sometimes I just can’t come up with the right word and so those of you in the front row, if you know the word I’m looking for, just tell me. You also noticed that I will speak to the floor and it’s not that I have become shy all of a sudden but if I look at you guys, you will distract me a little bit and by focusing on the floor I can sometimes find it easier to find the words. So I apologize that I’m that way for a while more.
The biggest puzzle that I brought with me, there are 2 of them.
The first one is that I wonder why success is so hard to sustain.
Now if you look across to sweep a business history, most of the companies which at one point were widely regarded as unassailably successful. A decade or two later you will see them in the middle of the pack and often at the bottom of the heap and the strange answer to that question was that I decided it was actually the principles of good management that we teach at places like Harvard that sow the seeds of every successful companies ultimate demise. In other words if you do everything right as we teach you, you will fail [Laughter]. I’m gonna talk about a little bit about this morning.
And the other puzzle that I brought with me is I wonder if the enterprise of creating new successful products and businesses truly is intrinsically a crap shoot?
The way you can frame the problem is the venture capitalist try to pull into their firms some of the smartest people in the world. People bring to them the proposals that they ask for funds. They put together a team of people that study this thing out the gazoo and they model it and stress it in every way they can and get smartest people that they can find to come in and understand it and put their holy water on it or not you know, and then when they finally decide it that this thing is gonna succeed they put they put their money in and it turns out that they are right one or two out of ten times. And so if ostensibly the smartest people in the world throw their guts into this thing and they can only be right 10 to 20 per cent of the time. Is the creation of new businesses it crap shoot and I decided in this case its actually not a crap shoot. There isn’t a cook book yet that there are a few things that you just do you find the probability of success is much than what most of us think given history.
So that’s the hope that you guys will have at the end of my piece this morning is that I really do think that success can be sustained and the career of building new businesses isn’t as problematic as historically we have thought. I will try to pause a few points in the time for you to throw cannon balls or criticisms or questions in my way.
Now what I wanted to describe here, first of all by the ellipses, is that you can almost diagram the history of every industry in this way in one sort or another. The one in the middle is for where most industries start. I have put it there because in most histories the early products are very complicated and very expensive, meaning that only few people who have skill and money can only use the product and then the larger ellipses are meant to represent people who have not as much money and not as much skill but when you can make products that we can sell to them it greatly grows the market. And the ones that started in the middle almost never drive it to the periphery where many more people have access to it, and they get killed.
So let me describe this model that we call disruption that emerged from my questioning. So a plot on the vertical axis is product or service over time and in every market there are 2 trajectories. The first one is there is an ability to utilize improvement. Now there is always a distribution of customers of the high end, very demanding customers who have very complicated problems they will never be satisfied with the best you can give them and at the bottom of every market there are simple folks with simple problems. You can over serve them with very little. So that’s the first element is in every market there is an ability to utilize improvement.
The second is in every market there is a different trajectory of improvement that innovating companies provide is that they keep introducing better and better products. The most important finding about this is that the trajectory of performance improvement almost always out strips the ability of customers to use the improvement. A few ways to visualize is to go back to mid-1980 when we were first learning to do a word processing on our early personal computers. Remember how often we had to stop our fingers to let the Intel 286 chip catch up. Because the world’s fastest processor couldn’t keep pace with our fingers but as Intel kept introducing faster and faster chips. But a few years ago when they introduced first 3 GHz Pentium 4 processor they had shot way beyond the speed that most customers in mainstream applications could use. And there are few people in the 4th standard deviation who still need faster chips. Now some of the innovations that help companies move up that trajectory are just incremental year to year improvements. Others have more dramatic breakthroughs. For example in telecom the transition from analog to digital and digital to optical were dramatic breakthrough improvements that cost billions of dollars. But it turns out that the purpose of the breakthroughs is the same as the incremental ones in that all they are trying to do is to sustain the trajectory of improvement on which they compete. What we found is that almost invariably the companies, the leaders in their industries on the left hand side before these battles of sustaining innovation began find themselves still on top of the industry when these battles of sustaining innovation come to an end and it just doesn’t matter technologically how difficult it was. If the purpose allows the leaders in an industry make better products and make better profits to their best customers. Always the leaders win.
But there is another type of technology that invariably kills the leaders. We call this one a disruptive innovation. We use the world disruptive not because it was a breakthrough improvement but instead of sustaining that trajectory it disrupted it by bringing to the market the product or service that is so much more affordable and accessible that a whole new population of customers now can own it and use it and what we found is that invariably it was an entrant company that came in and killed the leader and then disruption occurred. Those of you who remember a bit of this in data processing there was a company called Collabnet here in the Boston areas that dealt data software that was higher hierarchical and they work on mainframe computers and then a crummy little company on the other coast called Oracle came around and they introduced relational based technology that according to the standards the purpose for which Collabnet made its products the relational based software was not nearly as good and then it got better and better and better here in the middle competition and pulled all the customers in the middle into this one and then the larger circle salesforce.com is doing the same thing and that’s the mechanism by which the leaders get killed. Now what I like to do is just illustrate the dilemma this confronts us by going back in the history of computing.
Those of you who have little bit of grey hair might remember that there was a company here in the Boston area called Digital Equipment Corporation and Digital Equipment faring through almost all of the ‘70s and most of the’80s was probably the most widely admired of all the companies in the world. And whenever you read explanations about why they were so successful, inevitably it was attributed to the brilliance of their management team. Then about 1988 Digital Equipment fell of the cliff and began to unravel very quickly. When you then read explanations about why it had stumbled so badly it was always attributed to the ineptitude of the management team, same folks running the company. Well for a while the way I framed it was “geez how could smart people get so stupid so fast”. And that is the way most people accept the demise of Digital Equipment’s is that somehow the management team that had its act together at one point was out of its league at another. But the reason why the stupid manager hypothesis just didn’t feel right is that every minicomputer company in the world collapsed in unison. It wasn’t just Digital but Data General, Prime, Hewlett Packard, Honeywell. And these guys you would expect to collude on price occasionally but to collude to collapse was a stretch [Laughter]. And there just have to be something fundamental going on and it turned out that this simple model was quiet helpful.
So for those of you who don’t remember this, Digital made class of computers that we call minicomputers, they were about this big. We call them mini because they were smaller than the mainframes. They sold for about 250K and the selling process involved lot of training, support, service and software. You had to have the cost like that in the business to play in the game. Given that kind of cost structure they had to generate gross margins of 45% and the products sold for $250,000 and that’s how they made their money. Now in their environment every year people were coming in to see their management with ideas to make the next generation products. Some of the things that they were thinking about entailed going to the very high end of the blue trajectory, making better computers that Digital had ever made before. In fact these would be so good that you could start to do on that small platform things that previously had to be run on a mainframe computer. If you looked at those businesses they promised gross margins of 60% and you can sell them for twice as much. Now while the management was trying to decide if they should invest in those products they were other people saying, “Guys you don’t get it would you please open up the window and look out. Everyone is buying personal computers. Are you nuts?” And they would look out and indeed they could see that everybody in the ‘80s were buying personal computers. But they could see a couple of other things.
First one is, remember how crummy those early computers were? In fact Apple sold the Apple II as a toy to children. Not a single one of Digital’s customers could even use a personal computers for the first 10 years their PC was in the market and that meant that the more carefully they listened to the customers they got no signal from their customers that their personal computer mattered because in fact it didn’t to them. And then when they looked at the business plans they looked a lot worse because in the best years they promised gross margins of 40% they were headed to 20% quickly and you could only earn those paltry percentages on machines that you could sell for $2000. And so really the choice the management had to make was “Geez wonder if ought to make better products that we could sell for better profits to our best customers. Alternatively, maybe we ought to make worst products that none of our customers would buy that would wreck our margins. What should we do?” [Laughter] And it really is a dilemma and these principles of good management that we teach that you should always listen to your customer and always focus investments where profitability is most attractive. They provide very good guidance to move up that blue trajectory but then one of these things happen and again it’s a disruptive technology that makes what historically was complicated and expensive, this now makes it affordable and simple so that a larger population have access. Then those principles of good management makes it almost impossible for rational people to go after and that’s the innovator’s dilemma that emerged from my work.
Can we just think about where else the economy over the last 30 or 40 years this happened where somebody made something that incubated at the bottom or simple end of the market and then moved up to the point that it killed the leaders?
Prof. Clayton Christensen: I am sorry, yeah that’s absolutely in the process of making it happen isn’t it. Where else do you see it?
Prof. Clayton Christensen: Yeah yeah so the whole movement has gone through but Blockbuster got knocked off by Netflix and Netflix is trying to move up market and knock off the cables and some others.
Audience: Wikipedia has gone and destroyed Britannica.
Prof. Clayton Christensen: Wikipedia knocked those guys off. Absolutely right. Any of you heard of a company called Toyota. [Commotion] Remember how they did it? So they didn’t come in with Lexus’ did they? They came in with rusty little sub compact in the ‘60s called a Corona. That’s an America brand and then they went from Corona to Tsusho, Corolla, Camry, Avalon, 4runner, Sequoia and then a Lexus. And general Motors and Ford were back there on the blue space making big cars for big people and they’d see Toyota coming at them from the bottom and they would say “we gotta get those buggers.” And they would send down a Chevette or a Pinto [Laughter]. But then they would compare the profitability of doing a simple product like that versus the profitability of making even bigger products for even bigger people and it just didn’t make sense to defend the least profitable part of the business when it had the option to make more money and now the game is over for Detroit. Who is killing Toyota? Hyundai and Kia, the Koreans have stolen the next wave at the bottom away from Toyota and it’s not because Toyota is asleep at the switch. Why would they ever invest to defend the least profitable part of the business when they have the privilege of competing against a Mercedes in the luxury end of the market? And then Chevy is coming next from China and seriously none of us have to work about worrying about them at all. We could go on all day, we could talk about how Autodesk started out on the simplest of the PC platforms could only do really simple things and then step by step moved up the market and just blew out the prior raters in that field.
Now, what I am talking about here I hope you can see is 2 sides of a coin. The first one is what’s the kind of innovation that could kill your company? And then the other side is “geez if I wanna start a new company that could kill a leader, how would I do it?”
And I wanna recount another story that was quiet useful for me in understanding again why the leaders in the middle, how you can predict what they will do. So, one of the biggest disruptions in the history of mankind actually was when the microprocessor disrupted the Vacuum tool and you gotta really have grey hair here to remember this. But until the mid-1960s almost all consumer electronic products were made with vacuum tubes. They were about the size of a child’s fist. In a television they were about 30 vacuum tubes. That meant the televisions have to be like this costs about £2500 in today’s money and that meant that only the biggest people who had the biggest apartment and the biggest bank accounts could own one. The transistor came along and it was disruptive because it couldn’t handle the power required to be used in the market that existed. Every one of the Vacuum tube companies took a license to the transistor firm Bell Laboratories and they took their license into their own labs and framed it as a technological deficiency. In other words the transistor just isn’t good enough yet to be used in the markets that exist and as a group they spent in today’s money about 3 billion dollars through the ‘50s and ‘60s trying to make solid state electronics good enough that it could be used in the market and these are the giants of the industry RCA, General Electric, Westing House, Zenith. While they were trying to make the technology good enough out here competing against non-consumption meaning they are trying to make it affordable enough that a whole new population can access to it. The first application was a hearing aid in 1951 cause you couldn’t make this thing with vacuum tubes. And then in 1955 Sony introduced the first portable radio, 10 transistors, $2, fit into your pocket and boy it was a crummy product. It had Static field and a very tinny sound to it. My brother and I were raised in Salt Lake City. We saved our savings of dollar a piece and we bought one of these things. And we had to stand west to the Great Salt Lake to get reception but we were thrilled with this crummy product because it was infinitely better than nothing which was our other alternative [Laughter]. And it allowed us to do things that were impossible like we could sneak out and listen to rock and roll out of the ear shot of our parents. Had Sony tried to sell its crummy radio to the parents who had a nice RCA in the dining room it would have judged to be crummy. Then in 1959 Sony introduced the world’s first portable television. Again a very simple product but by making it affordable and simple a whole new population of people could now have one because it was infinitely better than nothing. They were delighted with the simple product. And so a booming new market emerged out here in this new market and the vacuum tube companies in the middle felt no pain. And then by the late 1960s Solid state electronics started to be good enough that you can make big products with it. In the next 5 years or so all the customers got sucked out of this center from the vacuum tube center of competition into this new one and every one of the vacuum tube companies vaporized. And it’s a very sad story because it’s not that the vacuum tube companies lacked vision. RCA saw the technology long before Sony did. It’s not that they lacked commitment. The guys in the middle spent easily 30 times more money than trying to make the technology good enough that it could be used in their market. But that was the rap.
Their framing was that it would be used in the existing market as opposed to use in the emerging market that they enabled. And you could just see how in order for this new technology to be used in the back market, Solid state electronics had to be more effective and more cost effective than the existing technology and during the ‘50s and ‘60s that was the very onerous hurdle that RCA had to surmount and Sony because they were competing against non-consumption, they made products that were better than nothing and it’s all they had to do and little by little it got better and better. And that’s the mechanism by which this happens.
Learn how great companies are run
At BoS we run events and publish content that is highly valued by anyone trying to build, run, and scale a great software company.
Sign up for a regular dose of actionable and useful content:
I just say one other thing about this, that is when this occurs, I point this at the vertical axis, it always has a different measure of a performance and so people in the back look at this and say its crummy and the people in the front don’t care because they measure it in a different way. I wanna talk a little bit about online learning which is disrupting the Harvard business school although nobody at Harvard would say it is but it is and the reason why it’s in the back where the traditional universities play the way you measure weather Harvard is a better school than Cornell and that’s the better than Utah state is the quality of their faculty which is measured by their research and what schools they get their PhDs and somebody who doesn’t have the same pedigree as Harvard faculty are judged to be not as good. But out here competing against non-consumption the way they measure performance is teaching oddly [Laughter]. And I don’t know if you guys ever take courses online but Holy Cow, this is good and getting better even while we are back doing all of our research. The University of Phoenix which some people would describe it’s the word it’s almost like apathy. Sony was the same incidentally. Those guys, University of Phoenix are spending $200 million every year making software to be better in teaching and I guarantee you Harvard spends this year about 0 trying to make teaching better. Anyway that’s enough. This is the things that I want you to take away, one) worry about the bottom thinking about who could kill you and if you wanna kill somebody get him at the bottom because if you come at the bottom you are like a little boy and you wanna kill a giant. You don’t want to ever head them where it matters to them but you wanna pick a fight where a giant is motivated to flee rather than fight you. And then the second one is when you think about how big is the market? It would appear as if the blue is the biggest market but in reality almost always if you make something affordable and simple you find that market is much larger even though at the beginning it looks like that’s not true.
I will just say this, there is a lot of stuff being hoped for in green energy. Let’s just takes solar electricity is one example. The US and European government so far have spent about $ 30 billion trying to make solar electricity good enough that we could use solar electricity in America and Europe. And it’s about 7 years away from being cost competitive versus power over the grid in Europe and America, 7 years away. It’s been 7 years away for 45 years [Laughter] and it still is because the existing system is getting better and most people are trying to beat them on a sustaining innovation. It’s just a pipe drain and it will never happen. If we tackle green energy in the context of America and Western Europe because the existing products are so good that to beat them is very difficult. And while I was worrying about this, our daughter Annie got sent by our church to go to Mongolia as a missionary and she worked there for a couple of years and when she was done she invited my wife and me to come and she gave us a guided tour of Mongolia and in the capital of Ulaanbaatar she took us to this massive open air bazaar and we happened upon a set of vendors that were selling dirt cheap solar panels. It wasn’t created out of silicon wafers but just glass with thin film on glass and they were packaged with small 6 inch black and white televisions and antennas and these things were going out of the market like you can’t believe. Well it turns out that about more than half of all Mongolians don’t have access to power over the grid and therefore solar energy is better than nothing. Who cares if it was cloudy today, I couldn’t have TV yesterday and amongst the 2 billion people in South Asia and Africa who have no electricity, solar electricity is a booming market growing at about 50% per year without subsidy. My guess is whether it becomes viable in America isn’t through massive government grants to MIT but it’s just the folks competing against non-consumptions starting with the simple products and making it better and better and better and that’s the mechanism by which it will actually become viable in America, not by somehow expecting that we can… shoot that wasn’t important [Laughter].
The next one, there have been a few companies that have caught disruption and became the leaders in the old even while they were trying to continue to make the core business more and more profitable. But in every case they did it by setting up a completely different business model and giving the new business model the charge to kill the parent. That’s the only time they have ever succeeded. So again, if I stay in computing, IBM dominated the mainframe business, there were 9 other competitors. IBM was the only one that made it to the minicomputer disruption, the other 9 got killed and IBM did it by continuing to make their mainframes in Poughkeepsie but they set up a different business unit in Rochester, Minnesota and gave it the flexibility to make money with 45% margins, whereas the mainframes acquired 66% margins and they were about 9 companies that made mainframes and minicomputers and IBM was the only one that survived the transition to the personal computers. They did it by setting up a different business in Florida, that could make money at 25% margins. The other minicomputer companies like Digital got killed. Then IBM made the next transition into a services company and the model that you get in your mind is like in biological evolution; individual organisms don’t evolve, they are born they die but as mutants gain market share a population can evolve even though the individuals were thinning. And you kind of get the same sense that the competition here that a business unit wasn’t designed to evolve. It’s organized to do a particular thing very well but they have to be born and they die. But a corporation can evolve as it starts new business units and shut downs the old ones and that’s where the flexibility came from for IBM. But nobody else thought they ought to create new business units.
I will just go over this, I wanna come to the next one. Now this is an idea that we are trying to figure out is truly intrinsic innovation a crap shoot or is it more predictable? And it turned out that this one is really important. So, what we decided is that the customer is the wrong unit of analysis when you are trying to develop a new product and the way we have taught marketing is that you should always listen to your customers and we have concluded that that’s wrong because, here I am Clayton Christensen and I’ve got a bunch of attributes and characteristics and there is correlation between my having these characteristics and the propensity that I will buy this product versus that product. But my characteristics don’t cause me to buy anything. What causes me to buy something is that you know stuff happens to me every day. Jobs arise in my life that need to be done and I hire products to do the job for me. The causal mechanism is that I have got a job to do and that causes me to go out and pull something into my life that gets the job done. The conclusion that we have reached and this is not… it’s not an issue of words [Laughter] but it’s a substance that understanding the job is critical to succeed with a new product.
I will describe it with a silly story about Milkshakes. So this is one of the big national fast food restaurants and they were trying to goose up the sales of their milkshakes. I walked into one of their restaurants and there were the sandwiches and over here were the desserts and the milkshakes were the line item in the desert menu. And these guys were very sophisticated and they had a profile of the quintessential customer that likes each of those individual products. So they had a profile of the quintessential milkshake consumer and I read the milkshake characteristics and I thought “Holy cow this is Clayton Christensen right there” and so they invite people like me into conference room and so they say could you please tell us how we could improve the milkshake so you buy more of them. They get very clear feedback they would then improve the products in the ways that their customer said and had no impact on sales or profit whatsoever. So as we were thinking about this we realized that the unit of analysis is the job not the customer and so we stood in a restaurant one day for 18 hours and very careful data whenever somebody bought a milkshake; what were they buying, what time did they buy, what were they wearing, were they alone, did they buy other food with it , did they eat it at the restaurant or go off with it? And it turned out as we added it up at the end of the day that nearly half of the milkshakes were sold before 8 AM . People were always alone, it’s the only thing they bought and always got in the car and drove off with it. So to figure out what job they were doing, we came back the next day stood outside the restaurant so that we can confront these people as they are coming off with their milkshake and we say “look I see what you are just doing here but I gotta understand what job you were trying to do that caused you to come here and hire that milkshake?” and they struggle, and so to try to help to them we say, “think about the last time you were in the same situation needing to get the same job done but you didn’t come here to hire the milkshake what did you hire?” and it turned out they all had the same job to do in the morning and that is they had a long and boring drive to work and they just needed to do something while they drove to keep themselves occupied. One hand had to be on the wheel but geez somebody gave them another hand there wasn’t anything in it and I just needed to do something to do while I was driving. I wasn’t hungry yet but I knew I was gonna be hungry by 10’o clock so I just needed something that would sink down and stay there for the morning [Laughter]. Boy I never thought about it this way before but you know last Friday I hired a banana, take my word for it never hire bananas they are gone in 3 minutes, you are hungry by 7:30, if you promise you won’t tell my wife I hire donuts twice a week. They are better but they are not very good they crumble all over my clothes, they are gone too fast and gets to my fingers gooey. Yeah I hire bagels on occasion but geez so dry and tasteless, I have to steer with my knees while I put the cream cheese on it and if the phone rings we have got into trouble. If you look under the seat you will see a snickers bar wrapper that’s because I had snickers bar once but I felt so guilty I just have never done that again. But let me tell you when I come here and hire this milkshake, it is so viscous, I can’t even pour it out and it takes me 25 minutes to suck it up that thin little straw. Who cares what the ingredients are? I don’t, I just know that I am full for the morning and it fits right in my cup holder and it turns out that the milkshake does the job better than any other competitors and the competitors are not Burger King Milkshakes. But its bananas, donuts, bagels, snickers bar, coffee and so on and then it turned out in the later afternoon and evening it was hired for a fundamentally different job primarily by fathers who have been saying “no” to their kids all week long and there have just been something they can say yes to so that their kids will think of them as kind parents. And so I’m standing at the counter and I order my meal and then my son, Spence orders his meal and then he looks up at me and he says, “Dad could you buy me a milkshake?” and I put my hand on his arm and I say, “Spence I would love to give you a Milkshake” and I say that it has nothing to do with Spence but I just want to feel good about myself you know and so watch what happens there its consumed in the restaurant with other food and with other people and I finish my meal and Spence finishes his meal and then the kid picks up that crummy milkshake and lemme tell you it is so viscous, it takes the kid forever to suck it up that thin little straw. You know, and I wait patiently for a while and then I wait impatiently for a while [Laughter] and they I say, “look Spence we can’t stay here all night” so we throw it away half consumed. Then they invite me as a customer who buys milkshakes and say, “so Clay how can we improve our milkshakes so that you buy more of them?” What am I gonna say because I hire for 2 fundamentally different products and then when they combine my response with all of the other 45-65 year old male slobs with children and they give one size fixed in product that doesn’t do any of the job so that it’s been hired to do.
I hope you can see how if you understand the job then you understand how to improve it so that it does the job even better and it turns out that they were improving the milkshake on a trajectory performance that was irrelevant to the jobs for which it was hired. So for the morning, how would you improve it? Well you wanna make it even more viscous right to take longer to suck it up, you stir in tiny chunks of fruit but not to make it healthy because they don’t hire it to become healthy but to just bring variety and unpredictability to a boring commute. On occasion it would go [pop] and it would wake me up and then you would move the dispenser machine from behind the counter to the front of the counter and give people a swipe card so they could just rush in gas up and go and never get caught I the line and the other afternoon job would be a totally different concept you know. But I hope that you can see that how understanding the job to be done is really critical.
Peter Drucker who is a lot smarter than me saw the same thing before I did when he said “The customer rarely buys what the company thinks it is selling them” and I would just say without any background of substance that the probability that a software product will be successful is somewhere near zero, if the concept was developed in a company that has decided what the customer wants and you develop a product and then you find they don’t buy it and so then you have to hire a, I don’t know if you have heard it but they call them marketers, and the reason you would have to have a marketer is you’re trying to convince the customer that they need to buy the product that you have decided they need. If instead you understand the job the customer is trying to do, you actually don’t need much marketing because customers would pull it into their lives and almost always those companies had somebody who is on the other side who knew the job and its understanding the job that’s critical insight in short supply. It’s not an ability to make products. Mark, this might be profound we will see but can you speak that way so they can hear you?
Audience: Does this explain Apple?
Prof. Clayton Christensen: Yeah, does this explain Apple? It actually does to a remarkable degree. I don’t think that jobs brought unique technological ability to the enterprise but it just watched what people were trying to do and could think about how could we help people with what they are trying to do even better and I think that was the key insight. And, I hope that, I don’t mean to say this in a self-serving way but apparently in his auto biography that’s gonna come out he only cited one business book that he said had an important influence on him and it was the one that I wrote and I think this is why it isn’t that I wrote a great book but is an important idea. Other thoughts? Yes.
Audience: Do you think Google is making a mistake creating or shutting down their labs?
Prof. Clayton Christensen: Do I think Google is making a mistake creating or shutting down their labs? I dunno how to say this, I think it’s a huge mistake but what I worry about is that the generation of new products comes from marketing people who are thinking that they are trying to understand the customer wants and it’s not connected with what the people in the labs were working on. If instead they take the, I mean some of the nicest people I know are marketers so I don’t wanna say bad things about them, but if you get the job nailed you actually don’t need marketing. Customers will pull it into their lives, talk about it to their friends and so the people in the labs need to be out in the world watching what people are trying to do and do it better, and oh my gosh I bet they would come up with so many more great ideas that connect, Whereas what you see is failure after failure and the problem is not in the labs in my view. Yes.
Audience: It seems that one of the challenges for big companies in investing in new technology is that investors are pushing them to invest in their current business model. How do you see that playing out?
Prof. Clayton Christensen: His point is that in a big company that is public, how do you deal with the pressure of the for short term results? There are two answers to that. One is that you have to begin investing in the next products long before the core businesses have become mature because if the core business remain healthy then you can invest in these new things and those guy won’t notice it. If you wait until the core businesses are mature and then you start it up their demand is really big really fast and that’s almost impossible in a disruption so it’s odd so you need to invest when you don’t need it. If you need it then your doomed and then the other if in fact you are doomed to failure then you should just leave the company and become professor with the Harvard Business [Laughter] because all you gotta do is talk about it not to do anything. Now there are four levels in a job, if you understand the basic job to be done then it allows you to say now what are all the experiences in purchasing the product, in using the product that we gotta provide to nail the job perfectly and very often we just think about the product and don’t think about from our point of view we just have to make the product, from the customers point of view though they gotta deal with this thing and dealing with it entails experiences. If you know what experiences you gotta provide then you gotta ask this question “ Oh gosh what have we got in integrating into the product and how do we integrate it so that we can provide the experiences to get the job done?” And just as a discipline that’s actually really important in a software product that downside is that it’s actually very easy to add more and more features and the more features you add often more complicated it becomes and this just provides a little bit of guidance and discipline to you and if you know what you have got to integrate then you can build a purpose brand and I will try to describe what I mean by that.
Where products are easy to copy, integration around a job. This I how we know what we have to integrate it an how we have to integrate it in order to provide the experiences to get the job done. It’s in the experiences we give and the way we integrate it that seems to be odd for competitors to copy. Jobs are very stable over a long period of time. They are not vulnerable to product life cycles.
Learn how great companies are run
At BoS we run events and publish content that is highly valued by anyone trying to build, run, and scale a great software company.
Sign up for a regular dose of actionable and useful content:
For example there is a job out there needing to be done, gotta get there from here to there with perfect certainty as fast as possible so it’s there whether or not you actually need that today, the job is there and if you find out one day you have that job to do it turns out Julius Caesar had that job to do and back then all he could hire to get the job done was a horsemen with a chariot, now we have FedEx and so the technology to get the job done changes overtime but the job itself is very stable over time. There is another job out there which is, ‘I don’t have time to read what I have ought to read but I need to come across to other people as being articulate and well informed’ and George Washington had this job to do and there weren’t people providing much to help him fake people that he was really informed and Clay Christensen had the same job to do and there are much better ways for me to pull into my lives the things to make it appear like I am intelligent but the job is very stable but that’s actually really important for marketing means is that the idea of product lifecycle is relevant In this world. Understanding the job is very stable.
We just need to understand it more and more and when you develop a product that does a job well the customers are quiet happy to pay you premium price. You notice when you download music people try to offer it for free and it was ClueGee and then Apple when they came in with the iPod they really got the job done. My gosh we were really happy to pay a dollar for a track. Another way to visualize is that there is a job out there that my son Mike found himself needing to get done last summer. He called me up one day and said “Dad I’m moving in my new house tomorrow and I gotta furnish the house tomorrow.” And when that job rises in your life the word Ikea pops into your mind and Ikea has been trying to roll itself out around the world for 40 years and nobody has copied Ikea, nobody. It’s not that they have secrets, it’s not that there is no money in it their owner is the third richest guy in the world and yet nobody has copied him and my sense is that their reason is that they have organized themselves around a job to be done, they provide the experiences and they are integrated the right way to nail that job perfectly and it has this brand which we call it as a purpose brand which pops into peoples’ mind when they have that job to do and I can go down all of those things most of them would say they have no competition. If you talk to people at SAS for example, they are one of their businesses is my darn there is nobody that’s gonna compromise the day they got their customers and they would tell you they don’t have competition because they’re organised in a different way. This is another idea that might help you look into the future. So if I wanna go back, the model of disruption, how do I kill other people and have them not kill me? The second one, competing against non-consumption is the way to create big new markets with simple products and then the third one is: If I try to listen to the customer it would mislead me, I need to understand the job. The next one I wanna just suggest that sometimes the way to catch a new wave is to change the basic business model. And I will apply this to products for online learning.
It turns out there are 3 types of business models in the whole world, the first we call is the solution shop business model.
A solution shop business is somebody who defines your problems and recommends solutions. So McKinsey and Company, I dunno if you have ever experienced them, they’re a solution shop business, you pay them a ton of money and they will come in and tell you what our problems are and recommend solutions. The activities that go in a hospital where they are trying to figure out what’s wrong with you is a solution shop type of business. Solution shop business get their money on fee-for-service basis and in healthcare people are really worried that all healthcare is paid on fee-for-service basis in most of America. It turns out that for this particular business that’s critical.
Then the second type is the crazy word but it’s a process related business where you stuff business in one ends and not complete broken and then you do stuff to it and then you shift it out to the other end. And so most manufacturing is like that but education is a process business. We bring in 900 incomplete people to the Harvard Business School every fall and then we do stuff to it and then we ship them out perfected to Wall Street at the end of 2 years. And that kind of a business, the business model is that they make money on fee for use I’m sorry, fee for outcome. And medical procedures are done after they have decided what’s wrong with you is a process based business where you can pay for result rather than end product.
Then the third one is a facilitated network like telecom, I send data to you and you send data to me and a lot of organizations are emerging to help patients take care of each other like the life of diabetes and so on, and education and software I think is historically has been in the middle and I think is coming on the right hand side. These guys make their money typically on a fee for membership and sometimes on a key per transaction, but it tends to move towards membership. What’s been doing historically is that the textbook business, somebody comes up with an idea to write a book then you try to convince the publisher to accept it. If they decide they will take it on then you print the books and then you got every district in America to adopt your text book and then and only then the teacher can teach it and it’s a process by which that’s done. Online costs are online software most companies have tried to replicate this business but do it online and so they’ll develop a class which is 9th grade algebra and by doing this the same way geez the only way those guys will accept 9th grade algebra online is if we can prove that its better than having the teacher teach 9th grade geometry online and that’s a very difficult thing to pull off when you are competing on a hit on basis back in the blue space earlier. But I think what will happen instead and those things cost hundreds of millions of dollars to develop the software. I think what will happen instead is that it will evolve into a facilitated network where parents who are trying to develop little tools for their children to get better at spelling, or teachers who develop products particular students who are way ahead or way behind, or students who develop these things for each other as facilitated entrepreneurs.
At the beginning these products will just be little tools that they can teach, use for each other to tutor each other. They don’t try to take on a course and then those things kind of add up into modules and the modules can then be combined into custom classes. But my guess is that those pieces will be interacted, exchanged in the network. That’s a very different business model from the one that we have today. But it’s really quiet exciting and if we can think about it in those words then sometimes if we are gonna take on an existing competitor and we gotta beat them in their market with their business model. Boy there are a lot failures but if we can think geez maybe we can have a very different business model then the barriers are not its owners. So that was the next thing I was gonna talk about. I have one more but any questions or comments on this any of the other ideas that we had. Back there.
Prof. Clayton Christensen: That’s a great question so, if you are starting a new stuff then the old ones can get pretty left over and bored or whatever… almost always. And this is not an exaggeration if they think about the concept of business they are in by the job to be done, what almost always happens is that they think that if you go back to the milkshake: so they thought they had overshot what customers could use as evidence by they kept improving the product and had no impact on what the customers bought. But once you understood the job to be done, “Oh my gosh what we thought was overshot actually not doing the job well enough” and It allowed them to kind of flip back to the left hand side and making integrated solution that nailed it perfectly and I bet that some doctoral student will come into our shop and go back, our studies of disruption. A lot of times when it had seemed that it had overshot in reality from a jobs point of view it wasn’t good enough and so that can bring in really quiet extraordinary insight and excitement into the traditional job. That’s a great question. Yes.
Audience: …Isn’t your concept that you should study ‘non-customers’?
Prof. Clayton Christensen: Yeah yeah there’s kinda two things that you gotta figure out by watching or living with them: What’s the job they are trying to do and then think about, what are they hiring to get the job done, sometimes there is all kinds of work around. This is actually a really funny story. So, Procter and Gamble has been using this a lot for a number of years not trying to understand the job but their customer and one of their people was sitting in an audience like this and looked at their seat mate’s suit and it was a dark suit and it was well polished and well creased, but man, did he smell bad and then they realized that you know sometimes you don’t need a dry clean, because whether it is dirty or not you can’t tell but you sure can tell when they smell bad and that’s how they came up with Fabreeze. So you just gotta watch what people around you are doing and figure out if there is cheaper way to get the same job done. That’s a nice question. One more and then I will go to this last piece.
Audience: The notion of jobs to be done could be used for continuous innovation as well…
Prof. Clayton Christensen: yeah that’s right, Yeah jobs to be done applies for everything an that’s a different question so disruption is how can I be sure I kill them rather than they kill me and that’s a separate and now we ought to understand disruptive is more sustaining, I just gotta be darn sure that the customers are going to pull it into their lives because you harness the causal mechanism behind the decision. That’s a great point Sir, thank you. Let me just, the Harvard Business School, we are getting disrupted by crummy low end on the job learning experiences like you are having right now [Laughter] [Clapping]. This is just a one thought that I wanted to that we are all living here and at some point we might invite vote for the next choice in presidential candidates and they keep reason the question that why don’t we have new jobs and it looks like that the economy seems to be getting better because of the profits keep getting put into banks but there are no jobs and as I worked on this problem, I just decided that the core reason is it truly is the finance professors at business schools, and I think I have out this fingers at most members of our faculty but this one is a big deal and the reason is that for whatever reason, our faculty have decided that we should measure profitability by ratios, and so you have internal rate of return and what internal rate of return does is it forces you to try to pull in the cash flow as close to the beginning as possible and so the more you focus on the internal rate of return the shorter the time horizon of the investors. And ROA the Return on net assets is a ratio and that causes you to try to get assets off the balance sheet. And you do that by outsourcing. Well, I could go through some of these others and it’s those measures at its core that’s causing us not to create jobs and I will describe how, I dunno if it’s true but these are just my thoughts.
There is one type of innovation that you might call efficiency innovation. How to make things at lower costs and generally efficiency innovations takes jobs out of the economy. What creates jobs and puts them into the economy are disruptions or you make something affordable and accessible so that a larger population can now have access to and when you make something that creates more people who are only able to use it then people have to hire more people to make it and that’s been historically, the causal mechanism behind the creation of true growth and over the last 15 years there are not nearly as many of these disruptions then there were previously, so the cloud of course is a huge disruption online learning. [Coughing] Sorry I got something. But they were not making as many as those as they used to and I think it’s because of the way we measure things.
So, net present value what do you do, well you foresee the cash flows from investing in an innovation and then you compare that with the assumption of what if we do nothing we compare it to what we are today and so the net present value and DCF implicitly in the math compare that cash flow versus the base case which is if we don’t do it but in reality if we don’t do it things get worse on a nonlinear trajectory and so by its very nature those measures underestimate the return of innovation of any sort. I will skip over this.
If you are outsourcing to get assets off the balance sheet, you liquidate the company and I will just go through this example between the Asus Tek which is a Taiwan company and Dell, and Dell at the high end still has a robust business but at the bottom in the consumer products they have been disrupted in this way. So, Asus Tek started by making this simple circuit board in a Dell computer and then Asus Tek came to them with an interesting value proposition. We have been doing the simple ones come to think of it you ought to be doing the mother boards because mother board technology is our competence not yours and if you let us do it we could do for 20% lower cost ad Dells analyst looks at it and realized, “Geez they could if we gave the mother boards not only can we reduce cost by 20% but we could get all of the circuit manufacturing assets off the balance sheet which is very capital intensive,” so they shoveled that over. Dells revenues were unaffected but their profitability improved as they got out of the mother board business, Asus Teks revenues improved and their profits improved as they got into the business. Then a couple of years later Asus Tek came out with an interesting value proposition. You know we have been doing the mother board come to think of this that the guts of the computers, you shouldn’t have to bother yourselves with the assembling of all the other junk because assembling isn’t your core competence, its ours and if we did it we could it for 20% lower cost and Dells analyst looked at it and realized “Geez they could and we had them do assembly not only could we reduce cost by 20% but we could get all the other manufacturing assets off the balance sheet.” So they shoveled that over. Dells revenues were unaffected but their profitability improved as they got out of the assembling and Asus Teks revenues and profits improved as they got into assembling. And then Asus Tek came back with an interesting value proposition, you know, dealing with all those crummy suppliers of components and working out all of logistics headaches and shipping the stupid computers to your dumb customers, managing the supply chain isn’t your core competence but its ours and if you let us do it we could do it for 20% less. And Dells analysts looked at it and realized. “Holy Cow they could and if we gave the supply chain not only could we reduce cost by 20% but we could get all of that current assets off the balance sheet” so they shoveled that over. Dells revenues were unaffected but now their profitability really starts to look good especially return on assets because they got no assets [Laughing] and Asus Teks revenues improved and their profitability improved as they got into value adding services. And then Asus Tek came back a couple of years later with an interesting value proposition. “You know the design of these dumb computers really is little more than component selection. We got all those relationships cold, there is no reason why you would have to bother to design the stupid computer and we could do it for 20% lower cost.” And Dells analyst looked at it and realized, “geez they could and we had them design it then we could fire all of our engineers, drive cost lower because our core competence is our brand after all.” And so they shoveled that over same thing result and then Asus Tek came back one more time but this time they didn’t come to Dell they came to BestBuy. You know I don’t understand why need to have those brands Compact and Dell and Hewlett Packard, we will give you our brand, your brand any brand at 20% lower cost and Bingo one is gone and the other takers place and I was able to use the word ‘stupid manager’ once because the least as we define it there is no stupidity involved on either side but in the end the first steps seems to be attractive but you need to think about where it ends and we do this all throughout the economy. We outsource IT to the Indians who are very capable, the auto companies outsource the tier 1 suppliers. The oil companies are becoming marketing companies heads and the pharmaceutical companies are doing the same thing, and the Wall street analysts are outsourcing their brains to Bloomberg which is a good thing overall.
I had a conversation you know, in the semi-conductor industry. Intel still makes its own chips but they are basically the only ones whose substance still makes the chips in America and everyone else gets them abroad and there is almost a badge for when you become a fabulous semi-conductor company, people admire you for having gotten all those assets off the balance sheet because you can be flexible and so on.
And I went over on a tour in Taiwan and China and this is the puzzle that I had: so it cost today about $10 billion to build a new fab and so you have the Americans over here with just this fixation that we can’t put assets on our balance sheet and so they wanna build companies without assets. And the people who do them are “sure we will build a new fab for $10 billion dollars and I am quite happy to leave it on the balance sheet” and so why did these wanna off when these people are quite happy to have it on and it’s not because we are short of capital, we are awash in capital. We are talking to one of the leaders as TSMC, I asked him this questions and he send “Yeah you are Americans measure profitability by ratios and by doing that you can compare one to another but there is a subtle problem with that as there are no banks that accept deposits to nominate in percentages” [Laughter] and he said, “there is actually another way to measure a profit and it’s not by percentages but the way we measure it is tons of money [Laughter]. And it actually doesn’t matter how much assets are employed because we generate tons of money.” And I think as consequence of measuring it in the way that we have, that we, it’s almost like this cocoon of finance people measuring profitability by INR and ROA, it’s created a massive capital in America that I would characterize as migratory capital and they are trying to find a company and put money into it and the minute the company is in they wanna get it out and they wanna get out pull more out than they put in and if they play that game well they think that they are generating value added and profitability. But that migratory capital is what causes us them to do these kinds of things, very short term time horizon and outsource stuff that could have been our core competence if we wanted them to. So I don’t know if this is true or not but I have been thinking about it and I actually think again that the business schools are right at the core of the problem so I’m sorry [Laughter].
A couple of questions and we can end. Yes. Is this gonna be a good question ? All of these people are waiting so Sp11: Something that’s closer to all of us as an example of what you were just saying, I wont tell the name of the company because its well-known by all of you and we are discussing this work we are a consulting software companie and we are discussing with them what their moto is and one of their measures is exactly contribution margin and as you said it makes all the sense by all the managers and everybody but everyone in the company is somehow measured by that and we are discussing how that whole thing works and the process and how they are selling and the whole thing. We just find out that they have in some of their products a target of 80% on their distribution and subsidiaries and that’s the ? margin and that everything’s just fine and then we start to look at what you are doing and what the customers they are dressing and then to identify that the potential customers where they would have to put some additional effort and we ask ok now you have the ability to do that why don’t you and they say because we are gonna be spending some more money to address those customers and that will give us just 50% contribution margin and we’ll drop our final margin and we have our bonuses on top of it.
Audience and I say ok I just totally understand that if you cannot address both folks are just going to go for the money. But the money you have, people have the ability to do it and they say we don’t wanna lower our margins so, they are just going for the percentage and not for the ton of money.
Prof. Clayton Christensen: Yeah it’s a great thing. I think the Brits and the Americans have this disease the most and the Germans are not afflicted by this much and the Asians do even better. One last question, Sir.
Prof. Clayton Christensen: Yeah question is spinning it off and Netflix how they get into that trouble. They really had 2 conflicting theories at work here. One is they had two business models inside of the company and doing ‘on demand’ or over the internet is a kind of disruptive business to the physical ‘get the DVD back’, the more you do this the less this kinda, so they wanted to separate those but on the other hand from a job to be done point of view you needed to have it all together because they compete against each other for the job and which one does the job better depends upon where I am and what time it is and all those kind of things. And I think that the challenge is how to get your cake and eat it too so, keep the thing coherent from a customer point of view and then keep them separate from the operating point of view so that this can grow and gradually that can die and I think it had to be either, or but this, I am not as smart as you know, we can see that retrospect but its taught us a lot from them, great question. You guys have been very patient so thanks for giving me this time. [Applause]
Learn how great companies are run
Sign up for a regular dose of actionable and useful content: