Eric von Hippel, a long-time affiliate with the Berkman Center, leads off our 2010 season of lunch talks with a discussion of “Household Sector” innovation. To explain his body of work, von Hippel explains that he’s tried to bring thinking about the communications space into the world of physical things, examining how processes we think of as affecting digital media can also apply to other forms of innovation.
Today’s talk introduces a national survey of innovation carried out by customers – there’s 2-3 times more innovation from consumers than there is from the industry. This counters our traditional thinking about innovation. We generally believe that manufacturers dominate innovation – users satisfy their own, personal needs, but manufacturers can spread costs across customers, allowing for innovation that serves wider audiences. As a result, our understanding of intellectual property tends to protect manufacturers, not users.
If we think about the emergence of markets over time, at the beginning of markets, there are very few users, which gives manufacturers little reason to innovate, and lots of reason for users to innovate. To illustrate, von Hippel introduces John Heysham Gibbon, the inventor of the heart and lung machine. Dr. Gibbon approached manufacturers to ask for a heart and lung machine. The manufacturer – quite reasonably – asked whether we understood the processes necessary to provide a heart/lung machine. And Gibbon, a physician, not an inventor, spent twenty years studying and documenting those processes. In the interim, the market for heart/lung machines matured, making it possible to match his discoveries to manufacturing processes at a point where it was profitable to manufacture these systems.
In many fields, the user is often the innovator – in the space of scientific instruments, 77% of the innovation comes from end users. Manufacturers don’t generally know this. That’s because user innovations don’t look like products to manufacturers. von Hippel shows us the first completely automated radioimmunoassay system, a hodgepodge of machines networked to serve the needs of a large cardiology department. When it was finally released as a product, it was a single, integrated system… and the manufacturer would happily tell you they’d invented the product – the user innovation was an idea, perhaps, but not a product.
von Hippel tells the story of Terry Fisher, the faculty director of Berkman. He’s a passionate ice climber, and added a leash to his ice axe, based on a poor experience he had hanging from an ice face – von Hippel points out that mortal peril tends to sharpen the inventor’s mind. Fisher wasn’t credited for the innovation and mentions that it might be hard to credit properly – does anyone really want a leash on their ice axe with Terry’s name on it?
To illustrate user innovation, von Hippel explains center-pivot irrigation to us. If you’ve flown over the western US, you’ve seen green circles dotting the landscape. This is a model of irrigation that’s far more efficient that setting up huge systems of piping – instead, fields are centered on a well and a rolling pipe rotates through the field. This technique was innovated by farmers, but is now in wide use. Ask the companies who manufacture these systems who invented them and they’ll claim creation. Show them a photo of the systems developed by farmers, and they’ll say, “But you should have seen their welds – they sucked.”
Carliss Baldwin and von Hippel offered a model last year that helps explain what spaces are more or less open to user innovation. This model considers communications cost and design costs that affect innovation. Single users care deeply about design cost – as an individual, a user can’t afford to innovate it if design cost is too high. John Gibbon’s innovation around the heart and lung machine may be the limits of possible innovation – at higher design costs, the user innovator is going to fail. If communications cost is quite low, and if you’re able to modularize the problem, you can allow users to build modules and take on much larger design cost projects: Linux. If communications costs were still high, as they were in the era of the heart-lung machine, these projects would need to be tackled by a firm, not by user innovation. We’re now at an exciting moment in time, where there are projects that may be more efficient for groups of individuals to tackle modularly than for firms to take on.
This year, von Hippel and collaborators have published a survey of the UK population that considers user innovation. They asked a sample of the UK population whether they created or modified a product in the last three years to “make it better for you”. 1.4% reported creating a product, 4.2% reported modifying a product, and 0.6% did both. That projects a population of 2.9 million UK innovators, often creating simple, novel products to address household needs. One reported putting a hook on the end of a fishing pole to allow the user to pull trees towards the ground and trim them. Another bridged a circuit within a washing machine to create a spin-only cycle. The mean investment in these innovations was 101£ and 2.8 days of time – the median was even more striking: 5£ and 2 days time.
The UK population is a “cake with raisins”. Most innovators work by themselves. But special interest user communities work quite differently. In these spaces, roughly 100% of innovators in the space collaborate with each other. Innovations interconnect and accumulate. As a result, these communities tend to host far more innovation activity. In whitewater kayaking, for instance, “users have developed everything” – they’re responsible for 73% of hardware innovation and 100% of the infrastructure (mapping rapids) was done by users, leading to a sport engaged in by millions.
Lead users innovate and often freely reveal their innovations. Some of these innovations appeal to other users. The user communities grow, and eventually user-founded firms enter the space. Finally, once market opportunities are clear, incumbent manufacturers enter.
User collaboratives can often out-compete producers in design because there are more of them, and because the space for innovation is open, which makes it very easy to adopt these innovations.
von Hippel closes with an observation in the power of intellectual property. A survey of 148 Dutch SMEs, studying product and process innovation, discovered that firms are very likely to patent their (internal, otherwise invisible) innovations and not share them, while individuals working in similar spaces usually didn’t seek patent production and often shared their processes. This may reveal basic biases in how individuals and firms behave.
As we move into discussion, von Hippel tells a funny story about business travelers and hotels. In the early days of the internet, travelers like me would take apart hotel phones so we could connect them to our computers and dial into AOL or to university servers. Hotels responded by putting tamper-proof screws on their phones. So travelers started carrying specialized screwdrivers… and hotels started putting phones in metal boxes. Some firms don’t like user innovation, even if users are pointing to a new market opportunity. von Hippel notes that it’s pretty weird that we expect manufacturers to innovate – we should expect them to manufacture efficiently, but perhaps to look to their users for design innovation.
Terry Fisher, who studies this space, offers the provocative suggestion that we may not need much intellectual property reform in this space, for the simple reason that user innovators rarely seek this protection. Perhaps instead we should work on finding ways to ensure that users are credited for their innovations rather than compensated for them.
Wendy Seltzer offers a counter-hypothesis: just when your user-innovator communities decide to commercialize their work, that’s when they run into a claimant who hasn’t built anything, but has already patented the innovation. That, in turn, can be a major obstacle to formation of these firms.