Wednesday, 26 July 2006

Eric von Hippel

Speaker: Eric von Hippel
Most widely known for his research on customer-driven innovation, Eric von Hippel is an economist and a professor at the MIT Sloan School of Management. He specialises in the nature and economics of distributed and open innovation.

He is best known for his work developing the concept of user innovation – that end-users, rather than manufacturers, are responsible for a large amount of new innovation. In order to describe this phenomenon, he introduced the term lead user in 1986. His work has applications in business strategy and open source software.

Hosted by:

Transcript

» Click here to view the video.

THIS IS AN EDITED TRANSCRIPT OF A BOSS CLUB ADDRESS GIVEN BY PROFESSOR ERIC VON HIPPEL OF THE SLOAN SCHOOL OF MANAGEMENT AT MIT. THE ADDRESS WAS GIVEN ON JULY 26, 2006

There's a lot to cover and let me jump in. Basically this has to do with something quite enormous and all I have is 45 minutes to tell you about it, so I better get going.

I should mention first that this is not only my invention. We now have a lot of people doing research on this, so the things that you see here are vetted by a number of people. So we are either all totally wrong or perhaps there is something to it!

So the basic idea that I'm going to be talking about is a huge shift. The traditional way to innovate has been - economists have thought this and companies have thought this - the traditional way to innovate has been to go and look at your target market, ask them what they want through marketing research or something like that.

It makes perfect sense if you are trying to sell something to somebody, you probably ask them what they want, and it is called find a need and fill it, and it also makes sense from an economic point of view because there's a lot of people there and you can do R and D around that large number of people.

But we began to study where innovations really came from and we saw something rather surprising and I will show you in a moment how we track them back, but what we saw when we tracked down the real breakthrough innovations in different fields, what we saw is that the functionally novel ones, the ones that were starting new product lines were done by the users. They were done at the leading edge of a field.

In the beginning there are some users who need something, but it is not a very big market at all. It's quite small. It is also uncertain, because even those users don't know quite what they want. So, for instance, a child who is really bored with going down the hill on roller skates might try to make a skateboard and in fact they did. What happens is they take apart their skates, they hammer them onto a board, they go down the hill, they say, "This is fun".

Now, they couldn't have characterised before to a market researcher what they wanted, but they were sort of co-inventing the sort of solution and the need as they went.

So, interestingly enough, this is true of consumer products and also, as you will see, of industrial products. So this now says that it is a very different kind of a mode that we have to adopt to get innovations into firms efficiently, because in fact marketing research looks at the centre, as I said, and regards these people as outliers, as systematically exclusive. So we are really losing what is going on here.

The average lag time between the time users innovate over here and begin to do that and the time that the first company begins to commercialise it is five to seven years. So by understanding what is going on we can collapse that and be quite more efficient.

What we did was we said: We want to identify over the past 30 years the most important innovations, the ones that really change things and we will track back their histories, and so we got these samples in a number of fields and then what we began to do was we actually tracked back where they came from, and we found two very interesting things.

First, we found that the manufacturers who produced them didn't really know where they came from. They would say, "Well, probably we invented them, if you're sending us a plaque". Well then, we would say, "Did you get any inputs from customers living or dead?" And they would say, "Well, yes, through marketing research", but you see, when we actually asked them the histories of these things, things got very obscure towards the front end. You know, they would tell us how they improved it and made it better and it was so great, but then when we got down to the core about where it came from, often these stories would begin with something like, "Well, one day a prototype fell out of the ceiling". So we thought: Ah, it's the ceilings. This is the key. But what it really was, was that users had innovated, users had innovated, other users began to pick it up and eventually, as I say, after five to seven years, what would happen would be people would start to hammer on the company or a serviceman would see it in the field, customers would say, "Why are we modifying this thing after we buy it from you?" And that's where it came from.

How does a user innovate? A user takes what's around.

This particular thing here is an automated radium (points to image) system. It was the first one. It was done by users at the University of Virginia. The reason they did it is because they had to do a heart study with 20,000 patients. The previous methods were manual. I know myself from my graduate students that you can get a graduate student to do a lot of things but you can't get them to do it 20,000 times in a row. They will begin to automate. So what they did was they automated. Now, they took what was around, right. They took pieces of things; they assembled it into something that would work. So from their perspective it was a product, and it in fact demonstrated the operating principles of this thing that the company later adopted, the first commercialising company. But by the time the company had made it really nice and made a spot in the corner that said Hewlett- Packard and made an operating manual, they began to think they had invented it, because this isn't a product, for God's sakes, right? The loss associated with this is not the issue of credit and who gets the credit. The loss associated with not knowing where your innovations come from is that you don't know where to find the next one.

So I am going to give you a couple more examples of user innovations. Once you start to see this pattern, you see them all over the place. We used to have Mount St Helens in the United States, remember, and it blew up. You recall this? This (points to image) is two days later, these are police cars and what you see here is that in fact what was going on was there was so much volcanic dust in the air that it was blocking the air intakes of the cars immediately, and so what they began to do is innovate. Now, you see the solution is not fully standard. It is only two days later but what you see is basically this is the kind of hose we use to attach our clothes dryers to the wall in the United States, you know, to get the hot clean air out, duct tape, you know, the universal anti solvent, and what they have done is they have constructed a big filter.

Anyway, now, why do you suppose that the users innovated? Why didn't they write to somebody who makes filters and say, "Oh, we'd like something like this"? Why didn't they do that? No time, right. They had to do it. So what you see at the front end of these things is that there are users who are experiencing need early and they need it enough to have to do something about it.

It is also true of consumers' goods. Mountain bikes and all these kinds of things were developed by users. I suppose you all have mountain bikes or had them. Anyway, mountain bikes over suspension systems and all the rest of it were developed by users. The manufacturers said you're not supposed to use a bike that way. So this poor kid would get a bike from his father and he would ride it down Mount Tamalpais and turn it into a pile of dust and he would bring it to the supplier and say, "It's broken", and they'd say, "What did you do?" And he said, "Well, I rode it down Mount Tamalpais". They'd say, "You're not supposed to do that. Your warranty's void", which was the standard approach. So the kids actually started putting motorcycle brakes on bikes and so on and so forth and building them in a very robust way. Now it is two thirds of the US market in bikes, probably yours too, and the manufacturers say, "You know, we did such a good job developing that".

An innovation is a user innovation when the developer expects to benefit from using it. It's a manufacture innovation when the developer expects to benefit from selling it. So this means that it can be an individual or a firm. Like Boeing, for example, when Boeing makes a new machine tool to help it make aircraft, that's a user innovation. They expect to benefit from using it. When you make aeroplanes it's a manufacture innovation. When you as an individual make something to use it, like an improved consumer product, that is a user.

So we have users innovating. The next step is: Which ones of these users do you want to get your innovations from? And the answer is you want to get them from lead users.

Now, lead users have two characteristics: One, they have needs that foreshadow general demand in the market place; two, they expect to obtain high benefit from a solution to their needs. So they're at the head of the curve and they have a strong enough need so they will innovate.

Now, if somebody innovates here, you don't care so much as a manufacturer, right. You want the ones that are ahead of a trend that is important to you. Now, the users that innovate don't care whether they're at the head of a trend. They're solving their own problem. You as a manufacturer care that they're at the head of a marketing trend. You don't want to just stand there and say, "Oh, users, if you have any ideas, come see me". What you want to do is say, "There is a strong trend in my market place. I will find the users that have an extreme need with respect to that trend today".

Great first solutions are often found in advanced analogue applications. Let's suppose that the issue is automobile braking, that is stopping, okay? Now, if you go to the middle of the market, here's a wonderful sedan, you'll see cars like that and users who buy cars like that and you say, "What do you want?" They say, "Well, I'd like to stop. It's always a good idea". But now, you move to the front, you decide that that's an important trend. You know, whether it's insurance companies or whatever it is, it's an important trend. Just like nowadays, this was an important trend ten or fifteen years ago with ABS braking. You know what that is, right? But now stability control and so on is a huge thing in automobiles to make them safer. So automobile braking, what do you do?

You go to the front and what do you find at the front there? That is people who have an extreme need for this thing and what you find is racecar drivers. Now, when you were kids you probably all learned how - do you have driver's education in Australia? You do, you don't just randomly go on the road, okay. I'm not commenting on your driving. You do very well. It's just on the wrong side of the road, but anyway, but what you were probably taught to do, is pump on the brakes like this. Now, ABS braking on the other hand is an automated version of that. That actually pumps the brakes for you.

Now (at airports) they cannot do what the automobile people do, which is put sand on the streets and salt on the streets to melt the ice because on an aircraft runway the salt would basically wreck the bodies of the aeroplanes and the sand would get held in the engines and damage the engines. So they had a different set of constraints. What they were forced to do is do something with the wheel itself, and so that's what they did. They came up with automated ABS.

Now, what I have told you is two things. First I have said users innovate. They are often the ones who actually develop the functionally novel new products. The second thing I have said is there is a way to find those innovating users that you want, because right now what's going on, the reason there's such a long lag, I told you, five to seven years, is the companies are not systematically looking for these innovating users. Instead, as I told you, they're looking at the centre of the market, finding some information, and then developing it themselves.

It is much more efficient to go out and find these lead user innovations because the lead user innovations contain the experiments that you would have to do if you wrote this thing and put it out as a prototype and watched the users use it. The users have gone through all that iteration for themselves. They have shown value in use and so on. You have a much better platform to build from. It is faster, more economical, more efficient. Okay, so I've said also there's a way to do that.

Now, the next thing I'm going to tell you about is that users are different. Basically I told you that users develop functionally novel innovations. Manufacturers tend to develop innovations too but they develop dimension of merit improvements, and the reason for this is that information is difficult to transfer, so each innovator uses the information it has. Users tend to have information, need information. Manufacturers tend to have information on a solution approach in which they specialise. So what you tend to find, for example, is the users inventing the skateboard or inventing the first kind of D ram, or the first kind of semiconductor processing machine and then what you find is that the improvement, the next generation D rams are done by manufacturers. This is why by the way, user innovation is so important, because it is the first of things.

All these users out there, you can't just choose any user and say, "So innovate. I'm watching", right, because what happens is that this combination of information that a user has differs from user to user.

Now (points to image), this drinking system. It used to be called canteens. It's a hydration system, right. Well, it was invented - what it is, is a bag of water, as you know, and a tube that you can drink from, instead of the way you did on a bike or something like that, you used to take a little bottle out of a bracket and drink it and put it back in. It was done by a guy who did summer bike races, so he had a need, and he was a paramedic. So what he had in his truck was IV bags and the tubing that you use to attach them to the patient, and what he did was he fashioned a drinking system from that tubing and an IV bag that he sewed to his shirt. Isn't that cool?

Now, the point of it is that innovation from users is distributed out there because we all know different things. So it is not like there is a simple user who will invent everything. It is a distributed phenomenon, because it's that happy combination of need of solution in a particular individual or firm that produces the innovation that's exciting.

The next thing that happens is that users share this information in communities. And the next thing that happens is that the really cool innovations that you can actually make a market with respect to, because a lot of people want it, bubble to the top. So not only do you have users innovating, but you have users selecting what will be successful.

Now the next scoop then is that user innovation is not rare. The reason these people form communities is because there's a bunch of them. So, 22 per cent of surgeons modify equipment, not all surgeons, university surgeons who in fact are doing leading edge work. So it's not like you will find everywhere somebody modifying equipment - aren't you glad - as they operate. And in the same way mountain biking, 20 per cent of the people were innovating - but it was not that I would go to the bike racks of MIT and I would find 20 per cent of the bikes modified - because we don't have mountains around there - but if I went to where they are actually pushing the sport of mountain biking, I can just go to any community and I will find 20 per cent of the bikes modified. Now, this means that there's a lot of user innovation out there, a huge amount. It is just that companies are not mining it.

So they innovated collaboratively, they freely reveal. So we can talk about IP, but they freely reveal their innovations to each other. That's how they're built. It makes economic sense.

Okay, so free revealing makes economic sense and what is going on now is that these communities are becoming so powerful that they are actually supplanting innovation by manufacturers, and I'm going to give you a cool example.

I am giving you these user examples in sports and so on because it is just more interesting but the same thing happens when you are developing anaesthesiology machines, it is the same story. These are kites. Developed by users, they have a big kite on top. If you haven't seen it, they have a little board and the idea is that you jump high and you try to stay up and do tricks in the air.

Now, notice all those wires and stuff. There's a lot of control involved in one of these things and this all was developed by users, not manufacturers at all, all by users. The next thing that happened was that manufacturers entered. It became a $100 million market in a few years. There were a few firms out there. They each hired one or two engineers; they were introducing new stuff. Then a guy named Saul Griffith at MIT - he was in the aero-astra department. He said, "You know what, I want to share this stuff I'm doing", and so on and so forth. And so users developed and posted kite designs and they were often better than the designs by the kite manufacturers.

Now, what happened was that people who were joining, some significant fraction of them were aerodynamicists from NASA who were a hell of a lot better than the people the companies employed - the company cannot afford to hire people like that - and they brought along these tools, like this is a vortex analysis model. Okay, isn't that cool? Because slow speed aerodynamics turns out to be kind of complicated. So they began to analyse this stuff and they were building this stuff and then they were posting their designs, first their sketches. And then what happened was that they figured out that they could post them as CAD files and they could have them cut out by a sail loft, you know laser cutter in a sail lost for boats. They cut out the cloth with the laser cutter. So they would just download over the web, you would post them on this site, your CAD file. Anybody else who wanted to could download it, could cut it as local sail cloth, have it sewn together and in a few days you would have a hundred people out there saying, "Well, this is how it works for me" and so on and, "This is an improvement" and all the rest of that.

The volume of activity in innovation just blew the manufacturers away. So what the manufacturers began to do is they began to in fact produce user designs. So look what we've got here. First, there's civic users innovating, they produced important stuff. Now what is going on because of the Internet and so on is that users can actually drive manufacturers back out of the development process. What happens is you even have it in open source but it's also true in the development of semiconductors and so on and so forth. They're just driving semiconductor designs. They're just driving, the users are driving the manufacturers back into production model and making the tools.

This is all very cool, and it is driving manufacturers nuts. They don't know what to do about it and I'm going to give you an example of Lego and then I'm going to give you an example of user communities and then I will stop talking.

So this is a Lego model. Do you know what Lego mind storms is? Anyway, what it is is a robot kit that you can actually build out of Lego parts. So some of these Lego blocks have a little CPU in it, some have sensors and so on and you can string this stuff together. Lego took ten engineers seven years to design this thing so that it was perfect - perfect. Are you paying attention? Perfect. Okay, and they sold it.

Now, three weeks later, three weeks later 900 hackers were working on it. They formed a community spontaneously. They began to modify it. They cracked the code immediately. They modified it, they improved it and in three weeks they had something that was much better, and then they began to sort of put it in the face of Lego. One of the things these things do is they follow a track around. You know, you sort of have a track of tape and it has a sensor, so it can actually follow the track of tape, and they began to show that in fact the thing they'd developed in three weeks could do it two and a half times as fast as the thing that Lego spent seven years developing and so on.

The product was very successful but the customers weren't the ones they expected. 70 per cent of the customers were over age 18, craze among adults, Silicone Valley firms forced to ban Legos at work. Anyway, great success and Lego didn't know what to do. They just sat there. They didn't know what to do. Users are not supposed to do that.

It was really funny. A colleague of mine had an article that he wrote showing 40 per cent of the users innovating and we sent it off to a journal called Consumer Behaviour and they rejected it with a sneer and they said, "Users don't innovate" - no, "Consumers don't innovate. They consume." We said, "Yeah, but 40 per cent of them innovate". They said, "No, they don't".

Now, Lego has begun to figure it out. So what they have done now with their next version of mind storms, what they have done is they have started to adopt the innovations of their innovating users and put them into their product. They also set up sort of a web site where they have allowed users to post their innovations and so on and so forth. They are innovating with this and it makes a lot of sense because there are about 200 people internally in R and D in Lego and there are 20,000 outside.

Now, my own theory is that 20,000 is larger than 200. So you may find for your firm that there's a hell of a lot of innovation going on out there that you don't know about. Now, Lego got to know about it but these were all user-founded groups; they weren't Lego founded groups. They had broken themselves up into different kinds of functions. So we have people who are interested in trains and we have people who are interested in robots. They had done all this on their own and what Lego is learning how to do is build a relationship, which they are now doing pretty well.

One of the things we run at MIT is an innovation lab and we've had companies come in there, we've got six companies that we have working with us at one time and they'll come in and they'll say, "There are no users messing with our product. It's too boring" or whatever it is. We go to the web and find tonnes of people innovating on your product. It is just that they're not calling your 800 number to talk to you, you know, but they are out there. This is true. Financial firms, you know, where people are taking the information that the firm supplies them and modifying it and plugging it back into the system of the firm the way the firm expects, so the firm has no clue that this is going on. It is true of toys. It is just true all over. It is true of automobiles, as you know. You will buy an automobile. The next thing you will do is take it apart and modify it.

Okay, now, so what's going on now is that people are thinking through how to do this and so a company called Strada, for example, which is software company making statistical software, knows that it does not develop tests, the users develop new statistical tests. So it made a product that has basically tool kits - has tools in it so users can do this more easily on the Strada platform. Then what the users do is they post their new tests and Strada watches how many other users download them and what's interesting to the user community and what's not, and when it sees something that's interesting it takes it and makes it part of its product.

Now, it does not steal it. What it does is it says, "This is Joe's test", and what some colleagues of mine have done is they've asked these various people who developed these, "What do you think about this? Are you getting ripped off or what do you think?" They say, "No, that's fantastic. I'm really glad that they're publicising my test." So they've found a way to be symbiotic about it. Now, what you have to do as we're evolving into this new system, where basically the division of labour between the users of something and the manufacturers is evolving, what we have to do is we have to sort of figure out how to do it effectively so nobody feels ripped off and it becomes symbiotic, okay?

Denmark now has decided to make user centred innovation national policy and we're participating in helping them, and while they're doing that, Australia ought to do it too and here's the reason: Basically what is going on is countries like the US and so on are doing technology push. In a sense what we're doing is funding R and D, and so we are into subsidies, we have all sorts of things like that, but it is hard for smaller nations to compete in that sort of R and D push race. What the Danes figured out is that, you know what, we can work on the technology pull side, I mean the need pull side, we can figure out a way to better interface with the user communities that are doing the innovation for us and thereby get a jump forward. So what we're doing with them is we have set up an innovation lab over there and we're working about developing training materials and we have some academics in place and so on and so forth.

This stuff by the way, if you want to study it or play with it, is really such fun or if you want to try it in your firm there's also stuff - I have a little company, www.leaduser.com and there is also free stuff that you can download from there.

Okay, so what I told you is that users innovate, that they produce some of the most important innovations out there, but marketing research and so on doesn't see them because they're outliers, that therefore the adoption track is a very long one. I have shown you that a lot of users innovate, they're forming communities and in fact what we have to figure out now, and luckily it's our generation's chance to do it, we have to figure out a new model of innovation that sort of builds on this kind of collaborative effort.

Okay, well, thank you very much.

Venue

Four Seasons Hotel