Article - Issue 17, October/November 2003

A missing piece of the innovation puzzle

Paul Baffes

Download the article (118 KB)

IBM’s Extreme Blue laboratory

In the midst of ever-evolving engineering technology, could a summer internship programme provide the link between technical and marketing innovation processes? Paul Baffes explains the value of inexperience as he uncovers the secrets of IBM’s Extreme Blue TM laboratory.

Innovation and survival

Innovation is the key to survival. We all know this at a basic, Darwinian level. What may not be as widely appreciated is that innovation is becoming increasingly critical for established businesses. This runs counter to the popular notion of innovation as the purview of small start-up companies. In fact, the larger the business the more important innovation has become.

The fundamental reason for this is the pace of technological change. For example, a few years ago, an individual building a new home might well have considered equipping each room with special outlets designed to carry a local area computing network. Today, such wiring is unnecessary: wireless networks have completely changed the way we link computing devices together. As a result, it is now quite feasible to consider buying a centuries-old manor and ‘wiring’ it for the information age without having to alter one square metre of the original structure.

The implication is that whole product lines can be wiped out overnight. Thus, for established businesses trying to show growth to investors, especially large businesses with many products or heavily engineering-based business, innovation becomes a key strategic issue. Perhaps it is the key strategic issue.

To some in the technical and engineering fields, this will not come as news. For many years now we have seen insightful academic work describing the impact of innovation on businesses. However, this in itself raises an interesting question: if businesses are already aware of this issue, particularly large corporations with considerable resources, why has innovation been so difficult to master? What is it about how we innovate that is not working?

There are many answers to these questions, including financing, management and the tendency in us all to fear the unknown. One piece of the puzzle, however, may lie in the model developed by the Extreme Blue programme at IBM. The Extreme Blue programme has developed a new process that helps corporations innovate faster; a ‘missing link’ in the evolution of an idea at its earliest stages. It is an innovation programme at IBM, but it did not start out that way. It innovated its way into becoming an innovation programme. Along the way, it discovered something entirely unexpected: the absolutely vital nature of diverse, inexperienced talent.

Traditional innovation: an accident waiting to happen

One of the reasons why innovation is hard is because it requires skills from multiple disciplines. For engineering-based firms, the two major discipline areas may be generally defined as ‘technical’ and ‘marketing’ (or ‘business’). Technical communities build new products and services and marketing communities determine how to sell those products and services to customers. Anyone experienced with bringing new products to market can confirm that these two communities are very different. They have different vocabularies, they approach the world with different tools, and they measure results in different ways.

Coordinating these disparate skill sets is neither easy nor straightforward. As a result, traditional corporate innovation has a built-in tendency towards what may be termed ‘an accident waiting to happen’ (see Figure 1). Most corporations are organised along functional lines that group together individuals of like disciplines: for example, technical development teams report up through technical management chains; and marketing organisations report to marketing management. Each group is funded and evaluated separately, and measured by different metrics. Most importantly, corporate processes tend to coincide with management structure; in other words, technical teams tend to work together, and separately, from marketing teams. This creates a tendency to build up a dual momentum that drives potential innovations towards a ‘crash zone’.

The standard failed innovation story goes something like this. Someone in research comes up with a brilliant idea for a new widget. Initial funding is found (probably by a forward-thinking technical manager) to begin to ‘flesh out’ the technical details. A small technical team is assembled which brings a great deal of excitement and enthusiasm to the fledgling project. This passion leads to promising early results and the team is gradually enlarged until it becomes a substantial investment. Somewhere along the way, what started out as an idea has become a movement with a life all its own but which may not be grounded in market demand.

Meanwhile, in another part of the organisation, someone from a strategy team has seen a market analysis report predicting explosive growth in a new market. Through diligent work and innovative thinking, this person shows how the company can dominate a niche in this market. This analysis, and the potential profits, are so compelling that a larger team is assembled to ‘flesh out’ the business case. Market research consulting firms are hired to perform customised studies which result in conclusions that support the original analysis. Perhaps even sales and advertisement campaigns are put in motion to capture ‘mind share’ for the technical offerings to come. Somewhere along the way, what began as an opportunity to make money has turned into an outlay of money preparing for a product that may never exist.

The crash zone emerges when one tries to unite these two activities and finds that the assumptions the two groups have made are incompatible. At this point the corporation realises a considerable amount of money has been spent and activity shifts from productivity to discussions of accountability.

Yet, interestingly, no one has done anything wrong. All of these activities, which may be collectively termed ‘innovation services’, are indispensable processes. Engineering firms, after all, do make things and one needs basic research to discover new things to make. Prototyping and other feasibility tests are crucial to proving technical approaches suggested by research. On the business side, secondary market research reports are critical to helping companies spot emerging trends. Follow-on primary research, customised to probe specific customers of interest, is essential for bringing out products that customers will actually buy. The fundamental problem here is one of separation; no matter how well coordinated, if innovation services are executed separately the crash zone is the likely ultimate destination.

The Extreme Blue missing link

A different innovation model that attempts to avoid the mismatches between technical and business innovation services is shown in Figure 2. This model has evolved over time in the Extreme Blue programme. The central idea is simple: force interaction between technical and business innovation services at a much earlier stage, creating a more robust result. Thus, rather than having two separate courses of innovation, each with its own momentum, the routes are redirected to a combined method where the realities of one process curb the assumptions of another process. Just as the combination of iron and carbon creates steel, this combination of innovation services yields an entirely new process that is stronger than its individual components.

The result is not merely a combination of existing innovation services, but rather a whole new kind of process. A technical innovation service, such as prototyping, requires many business assumptions even if the technical individuals involved are unaware of them. Consider, for example, the very technical task of adding features to a software program. Even at the roots of the program, assumptions must be made about how the software will interact with the ultimate end user. These decisions are as much business decisions as they are technical decisions. In a parallel sense, it is very difficult to perform constructive primary market research involving customers without showing them the actual device in question. For example, what customer, circa 1970, would have been able to answer useful questions about a potential mobile phone product without first having the opportunity to use the device? One simply cannot build something useful without customer input, just as one cannot expect good customer input without giving the customers hands-on experiences.

Figure 3 compares the Extreme Blue ‘BizTech’ process with traditional innovation services. The left of the diagram depicts the innovation services typically associated with technical disciplines: basic research and prototyping. In isolation, these result in products with built-in assumptions about the customer. On the right of the diagram is the analogous set of business processes which, when involving emerging technologies, must make assumptions about technical capabilities. The Extreme Blue method combines these techniques by literally combining people who have different skills. Small teams, typically consisting of four individuals, are formed and given a potential opportunity. The team always consists of technical members (e.g. programmers) and at least one business member (e.g. an MBA). The team also has mentors who act in the role of facilitator, helping the team gain access to resources it needs. The opportunity comes typically at a very early stage: at the beginnings of a technical idea along with only a partial understanding of the market opportunity. Nothing is predetermined to an absolute degree; it is the team’s job, and ultimately the team’s decision, to take the project in whatever direction the team decides is most appropriate, using whatever innovation services it deems necessary.

One final ingredient is crucial: the existence of a ‘final exam’. When the team is given its project, it is also given a time limit at which point it is expected to make a formal presentation at an Expo for senior executives of the corporation (in 2002, the UK Expo was made possible through the generous support of the Royal Academy of Engineering). At this presentation, the team must show some new value to the corporation: a new potential product, customer interest, or the validation or debunking of some key assumptions upon which further investment hinges. Note that it is perfectly acceptable for the team to show further investment should not be made. The effect of such a final exam is electric; it has a wonderful way of knocking down the normal barriers between individuals by presenting them with an opportunity to make a real difference in front of an audience with the authority to take real action. In 2002 alone, the Extreme Blue programme generated the following results: nearly one patent for every four interns, three donations to the open source community, and 11 new potential products and services.

The role of talent

The Extreme Blue programme initially began as a means for attracting top talent. IBM, like many large corporations during the boom years of the 1990s, had trouble attracting talent that was enamoured of the start-up culture. Nobody, it seemed, wanted to work for a large established firm. Excitement was to be had in the work-out-of-the-garage settings where it was believed true innovation was taking place.

In this context, IBM knew it had to try something different to attract top talent and the Extreme Blue programme was born as a high-profile summer internship programme. This had three profound effects that were initially obscure. The first was to set a time limit, corresponding to the limits of the summer break between academic seasons, which injected a sense of speed into the programme. The second was to create a mixed talent environment, due to the fact that the talent requirements of the corporation included needs for both business and technical skills. The third was to set a precedent for attracting ‘unblinded’ team members: individuals who are extremely bright and talented in their own right, but who cannot possibly be steeped in industry experience. The resulting forces combined to shape the programme into one that serves a particular niche in the quest for innovation: rapid, project-based, hypothesis testing executed by bright, unbiased talent guided by experienced industry mentors.

The greatest surprise of the programme has been the strategic value of the ‘unblinded’ team members (see Figure 4). A nearly universal reaction from mentors who participate in the Extreme Blue programme is surprise at the value ‘inexperienced’ people can create in a mere 12 weeks. As an example, consider the story of Emma Shepherd, shown here.

The Extreme Blue programme – past, present, and future

In 1999, IBM launched the Extreme Blue summer internship experience at only a single site, Cambridge Massachusetts, and with only 24 interns. All of the interns were technical students and the projects were selected to be technically challenging and appealing to students. From that first year, the Extreme Blue programme quickly learned the importance of adding business thinking into the mix and the need for expansion. In 2000, a second site was added in Silicon Valley at IBM’s Almaden site in San José, California. In 2001, EMEA joined in, followed by China in 2002. Today, nearly 350 participants, both university students and early-career IBM employees, join the Extreme Blue programme to work on nearly 60 projects worldwide and year-round.

The strategy continues to be a mix of talent and innovation. The Extreme Blue programme exposes a group of bright young technical and business students to the best that IBM has to offer. Interns have access to world-class, and worldwide, resources and an unparalleled opportunity to network with industry leaders and develop professional skills. The Extreme Blue full-time staff work closely with internal sponsoring organisations to design optimal project proposals that will test new hypotheses for the marketplace. In parallel, the Extreme Blue staff utilise best practices for recruiting candidates, thoroughly assessing the skills of student applicants to match potential interns with the goals of our sponsoring organisations. At the start of each internship cycle, the interns are presented with the project proposal and are then expected to take ownership of the project, modifying or even completely changing the project direction based upon their perspective. Many students express surprise and gratitude at the faith placed in their abilities; at the same time, many project mentors express surprise and delight at the insights and creativity of the interns. This symbiotic mix, this blending of experience and freshness, of younger and older, of diversity in race, gender, and backgrounds, all serves to create an open and frank atmosphere where the best ideas can emerge and be acted upon quickly. It has become a hallmark of the programme which has sustained it and continues to drive its growth within the company.


The lesson of the Extreme Blue programme is clear: traditional innovation processes are vulnerable to unconsciously embedded assumptions, particularly when innovation is in its earliest stages. More importantly, diverse thinking, embodied in diverse individuals trained with differing skills, is absolutely critical. Combining such talents early in the innovation process creates a new innovation method that can quickly validate, or invalidate, an approach. Today, strategic groups within IBM now bank on that diversity of talent as a vital sounding board for testing new ideas.


Many thanks to Becca Loader for her help editing and coordinating, and to Ann Strosaker for her assistance with diagrams and especially for her permission to reprint her rendering of the Extreme Blue model (Figure 4). Thanks also to the entire Extreme Blue staff and summer teams whose unflagging commitment made this article possible. IBM and Extreme Blue are trademarks of International Business Machines Corporation in the United States, other countries, or both.

A nearly universal reaction from mentors who participate in the Extreme Blue programme is surprise at the value ‘inexperienced’ people can create in a mere 12 weeks.

A virtual secretary

Emma Shepherd, a 2001 UK Extreme Blue alumna, and her team explored the validity of a ‘Voice Enabled Personal Assistant’, a virtual secretary that one can talk to in natural language, for example to read voice-mail or to set up a meeting. Emma’s key insight was to redirect the project away from proving whether the technology would work to testing whether the technology was workable. Emma built a website for dynamically changing the settings of the virtual assistant. This offered a solution to the inflexibility of many existing voice enabled applications by allowing users to set their own preferences such as the speed at which the assistant reads. By the end of the 2001 programme, Emma and her team had amply demonstrated to IBM that natural language can be extremely powerful but also very complex for everyday use, enabling IBM to take the decision not to pursue application of the technology at that time.

Further reading

For official information on innovation within the UK see:

Other excellent sites for innovation within the UK and EMEA are as follows:

Paul Baffes

Manager of Projects Extreme Blue Programme, IBM

Paul Baffes is the Manager of Projects for IBM’s Extreme Blue programme. In that position he is responsible for selecting projects that will have maximum innovative value for IBM. His previous experience spans the entire innovation process including research, invention, development, founding a company, fundraising, marketing, and sales. He holds multiple patents and was NASA’s 1995 Inventor of the Year.

[Top of the page]