Flip (25 page)

In chapter 5, 'There is No Wisdom in Crowds', I argued that you have to be open to ideas from the fringe to innovate. You have to be open to serendipity, to finding things you weren't looking for. The open-source model does this by recruiting the energies not of a proprietary research force whose resources are limited even in the largest companies and organisations, or even the most powerful governments, but of anyone who cares to contribute. This ensures that new ideas will have a chance to emerge from the fringe. It's a flip on a flip. The true wisdom of virtual crowds is that people will be pursuing their individual interests and agendas. It's a market of the mind, which winnows out the good ideas from the bad in an uncontrolled, Darwinian survival of the fittest.

Nothing demonstrates the potential of the open-source model better than Linux. Linux itself grew out of two earlier open-source projects – Unix and GNU.Unix, the free-time pet project of two AT&T Bell Labs researchers in the late 1960s, was initially quite successful but quickly became heavily protected by patents. In response, GNU emerged as an opensource alternative to Unix, but lacked some central elements of a viable operating system, such as a microkernel.

Linux, the brainchild of Finnish computer scientist Linus Torvalds, was released in the public domain in 1991 and was a fully viable operating system. It is hugely successful. IBM and other large IT corporations use Linux as an operating system, and there are numerous small companies selling Linux software, which runs on everything from supercomputers and servers to mobile phones and PDAs. In 2008 the global Linux market is projected to reach $35.7 billion a year.

None of this could have happened cost effectively within a single company's, or even a consortium's, proprietary software development. To compile the millions of lines of source code in Red Hat Linux – the first important, but now discontinued, product from Red Hat – would have taken 8000 employee years of conventional development time at a cost of $1.08 billion a year.

I said that open-source development represents the true wisdom of crowds, and the thousands of contributors to Linux certainly make up quite a crowd. But this does not actually contradict chapter 5's flip, 'There is No Wisdom in Crowds'. To appreciate this, we have to look closely at the development of the Linux kernel and consider the concept of
lead users,
people who innovate solutions for their own specific needs in advance of similar general needs.

Only 2 per cent of the current Linux kernel is directly attributable to Linus Torvalds. But by functioning as, in effect, a lead user of the Unix-like concepts and designs that he drew on to create the first Linux kernel, Torvalds put his stamp on everything Linux and pointed out a direction for its future development. By the same token, Lego's Mindstorms User Panel, recruited from outside the company, served as lead users for the development of Mindstorms NXT. By definition, lead users are to be found on the fringe of mainstream activities, or to put it another way, on the cutting edge of developing trends.

Lead users act like fringe ideas that attract mainstream interest. As I said in chapter 5, innovations don't begin in crowds, but crowds are necessary to validate innovations. There may be no 'I' in team, but there definitely is an 'I' in innovation, and lead users put it there. An article published in
BRW
here in Australia cited research that looked at fifty projects in development over about a year and found that products using the lead-user approach generated returns which were on average eight times greater than products developed the conventional way. In his book
Democratizing Innovation,
Eric von Hippel shows that lead-user product development can be a far more effective means of innovation than conventional product development within a closed system. At 3M, for example, von Hippel and his research associates found that lead-user development resulted in five major new product lines with annual sales forecasts of US$146 million per product, whereas conventional solelywithin- the-company product development over the study period resulted in forty-one incremental improvements to existing product lines and only one major new product line with an annual sales forecast of US$18 million.

Eric von Hippel says of the lead-user product development process, 'This is not traditional market research – asking customers what they want. This is identifying what your most advanced users are already doing and understanding what their innovations mean for the future of your business.'

Some of the most interesting new branding and product development today are being done by companies that use social networking technology to solicit contributions from lead users and/or validate innovations by seeing what customers in general think of them. For example, the Vancouver-based shoe designer John Fluevog, who has boutiques in several North American cities and distributes his shoes to high-end shops around the world, practises what he calls 'open-source footwear' by inviting design submissions from customers. Although only a few such submissions become actual products, Fluevog says: 'Even submissions we can't make add to the stimulation. Our customers get more involved, and we get insight into who they are and what they're doing.'

Consider a slightly different version of the same phenomenon – the InnovationXChange. I had the pleasure of meeting John Wolpert, who developed the BRIDGE methodology, which instructs professional intermediaries on how to help companies share intent and find mutually beneficial collaborations without exposing secrets to each other directly.

Less than four weeks later I serendipitously bumped into InnovationXChange CEO Grant Kearney boarding a plane to Sydney. IXC is a not-for-profit company which helps companies, universities and research institutions collaborate and share information in the midst of highly complex international intellectual property laws, while still maintaining strict confidentiality. In their own words, they 'make it safe for organisations to "talk".' They have dealt with everyone from food makers to pharmaceutical companies, in one case linking a pharmaceutical company working on a partially finished drug with a nanotechnology company on the other side of the world that revolutionised the drug's delivery mechanism. IXC is a fascinating organisation, adding real value to Australian businesses.

This trend is global. For instance, the website Global Ideas Bank (
www.globalideasbank.org
) is a public forum where people can post ideas at any level of development (from a small inkling of an idea to a well-progressed idea that is hitting stumbling blocks) and people can comment and contribute.

SOME THINGS ARE WORTH PAYING FOR

Innovation contests and prize challenges are another means of open-source, lead-user product development. The pharmaceutical giant Eli Lilly recently incubated and spun-off a new enterprise called InnoCentive, which acts as a middleman between companies ('seekers') and a worldwide community of self-selected scientists and engineers ('solvers'). Seeker companies post challenges with cash awards for the development of new products and processes, and scientist solvers post their suggested solutions by a specified deadline.

In early 2007, InnoCentive's open challenges included an award of US$140,000 for new pressure-sensitive adhesives and US$1,000,000 for a biomarker for the disease ALS (amyotrophic lateral sclerosis), as well as US$15,000 each for a novel method of dust control and packaging to limit the breakage of snack chips. Successful solutions to past challenges include a new method for assessing the risk of breast cancer, UV-resistant coatings and a next-generation paper binder, among several dozen others for companies such as Procter & Gamble, Du Pont, Dow and Boeing.

Contests to spark innovation have a long history. In the eighteenth century the British government offered £20,000, an enormous sum of money back then, for the solution to finding an accurate longitude on the open sea. Instead of any of the highly credentialed astronomers of the day, it was a clockmaker who found the solution. His name was John Harrison. After four different incarnations of his longitudinal chronometers he finally built one that worked extremely accurately. For petty political reasons he was initially refused his reward! He fought it and fought it, and finally –
years
after building a properly working device that could measure longitude at sea – he was granted his reward.He was eighty by then, and lived only three more years. Thankfully, today such contests are known for paying up more reliably. Can you imagine the PR and word-of-mouse frenzy that would follow if they did not?

The American DVDs-by-mail service Netflix, which operates on a lending library subscription model, asks customers to rate the movies they watch and then uses an algorithm to make recommendations of new movies based on over 1.6 billion total customer ratings. As with similar algorithms that lie behind Amazon.com's customer recommendations and Google's search-tailored advertising, the more accurate Netflix can make its 'Cinematch' algorithm the more customer satisfaction and loyalty it will engender and the better positioned it will be against competitors, including coming digital download services.

In October 2006, Netflix announced a US$1 million prize and a five-year submission deadline for an algorithm that would increase the accuracy of Cinematch in predicting customer responses to movies by at least 10 per cent. According to
The New York Times
, the company, which did not expect quick results, 'underestimated the power of an open competition'. Less than six months after the start of the competition, the model submitted by the leading contenders, a team of Hungarian scientists, was 'already 6.75 per cent better than Cinematch'.

Incidentally, showing that it knows how to 'Think AND, Not OR', in January of 2007 Netflix introduced a new feature to its DVD rental subscription plans. For no additional charge, Netflix customers got the capability to watch streaming movies from Netflix on their computers, as well as continue to receive and return DVDs through the mail.

On a side note, Netflix sounds a little bit like a nascent kind of web 3.0, but it probably falls a bit short. It is certainly an example of organising information. But to truly be a part of the semantic web, the data would have to be ordered and stored in such a way that machines could understand what the data
meant
so they could parse it for useful information. It would need what information specialists call a
meta-language.

Netflix really is just a good database with a good search algorithm. If it was really web 3.0, you should be able to ask the database, 'I want to see a funny film, but I want it to be a dark humour, not slapstick, and I want it to have a serious edge.' The only way of doing that with Netflix would be to wade through all the user comments. In web 3.0, the computer could understand qualitative judgements and comments.

Summing up the benefits of innovation contests,
The New York Times
noted the 'two essential features of prizes. They pay for nothing but performance, and they ensure that anyone with a good idea – not just the usual experts – can take a crack at a tough problem.'

Procter & Gamble adopted a strategy called Connect and Develop as part of its fight-back following its market slump in 2001. It employed a network of scouts to look for (connect) new ideas and then to develop them. Procter & Gamble have an internal R&D operation with a nearly US$2 billion a year budget, but they are smart enough to still 'Think AND, Not OR' and look outside too by using InnoCentive's 80,000 independent, self-selected 'solvers'. It's like an R&D version of e-lance.com, where the company posts R&D projects and people can contribute. Procter & Gamble's decision to move away from just traditional inhouse R&D and combine it with this new outsourced variety has led to a 60 per cent increase in R&D productivity and the launch of more than one hundred successful products, including the hit Olay Regenerist.

Cisco Systems have an outstanding internal culture and empowered staff, but another vital factor in their longstanding success is an 'acquire and develop' strategy: they buy up small, innovative companies and use their resources and market clout to realise their potential. Rather than create an open competition, Cisco will let fringe entities develop an idea on their own time and with their own money until that innovation looks ready to be integrated into existing technologies or commercialised in its own right. Cisco will then buy the company, ideas and talent both, and take it to market.

Or for another favourite, Intel have built 'lablets' in strategic universities to work closely with researchers and graduate students to publish research that Intel will not own. The goal is to build the amount of innovative activity that is being engaged in the areas that Intel works. In the future they could adopt a connect-and-develop or even an acquire-anddevelop approach to some of the innovations created from these relationships.

These strategies make more sense when you consider data published by the National Science Foundation in the US showing that between 1981 and 2001 the percentage of total R&D expenditure that can be attributed to large businesses has shrunk from 70.6 per cent to 39.4 per cent. This is especially telling because the percentage that can be attributed to businesses with fewer than one thousand people has grown from 4.4 per cent to 24.7 per cent. It makes sense to connect, acquire and develop when smaller firms are where a significant chunk of the action is.

The other advantage of these approaches by companies such as Procter & Gamble, Cisco and Intel is that these smaller entrepreneurial ventures are putting everything on the line and have already taken the risk. They are generally more nimble organisations, able to move quickly when the market changes, and in reality their success depends on getting it right. If a large company fails to innovate in one product area, there are a few thousand others to absorb the loss. If one of these smaller companies fails to stay on trend it will find itself out of business fast.

As a side note, of the four forces of change, compression of time is the most likely to prevent you from embracing the opportunities that open-source and leaduser development present. As you try to keep up with change, you are forced to move more quickly. The faster you move the less likely you are to take the time to find and evaluate external options. You need to get over this natural focus, and get out and see what is happening.