XII.1 January + February 2005
Page: 24
Digital Citation

The innovation pipeline: design collaborations between research and development

A. Henderson

back to top 

There are many great ideas sitting on the shelves of research laboratories. Frustrated researchers often carry these to other companies where they become products. Why is it so difficult for a company to get the innovation that is carried out by its research labs into the products its sells?

HCI is a special case of this problem. User-centered innovations, yielding concepts that would be of great value to users, never make it to product, or they get watered down, re-engineered, and washed out on the way to becoming unusable products.

Our point of view at Pitney Bowes is that HCI and research in general tend to look at the development of products much too narrowly. Specifically, concepts must meet not only the needs of users to become products; they must address the needs of the many other players that participate in turning a concept into a product.

We believe that development can be more effective if all these needs are considered concurrently. We have therefore restructured research at Pitney Bowes (the organization is named Advanced Concepts & Technology, AC&T) to include a technology-grounded "Innovation Pipeline." This pipeline implements a practice of collaborations with the Business Units (BUs) in conceiving and trialing concepts for opportunities that stretch the BU's and customers' thinking.

I offer this practice as a good way to understand the relationship between research and HCI. As long as HCI works alone, it will have little effect on products. The problem, as I see it, is that considering HCI issues introduces tensions into the development activity, forcing "trade-offs," which often lead to question of what value user-studies add to a product. If HCI participates in product development from the very beginning, however, and collaborates in the struggles with the tensions it introduces, it has the chance of profoundly shaping and directing the product throughout its development and deployment.

I suggest we broaden the scope of what has to be designed in making products, and bring HCI to bear as a collaborating participant in the evolution of solutions to the much tougher design problems thus created.

back to top  The Product Sweet Spot

If a product is to make money for a company, it has to simultaneously meet (at least) four interacting sets of needs (See Figure 1).

  • Technology: It must be technically feasible to build the product. This may require scientific and engineering breakthroughs or classical technological innovation (cool technology).
  • User value: The product must provide value to the user. This includes meeting user needs, through a usable human interface, and also enabling novel or more effective user work practices when the technology is put into use.
  • Business Value: The product's value to the customer must be something your company can capture and supply at a price that is less than the effective value to the user, and greater than your cost in creating it. That is, it must make just plain good business sense to both your customer and you. This requires developing a value proposition and a business case.
  • Strategic value: It must be coherent with company strategy and be designed to advance that strategy. This may require restructuring company thinking.

What makes this really difficult is how these four value sets interact. For example, the need to adjust the cost to make a product (the classic FAC: fully allocated cost) is affected by all four of these concerns: cheaper parts and/or a different architecture (technological feasibility); a simpler user interface or decreased functionality (user value); increased functionality (business value); an entry into a new market (strategic value). Some of these pull one way, some another.

The job for development is to address all these concerns at once: that is, to get into the sweet spot. As with all design, this means managing the full collection throughout the development process and evolving a solution that satisfies everyone. While isolated pieces of work can be done on particular issues, none can be considered final until all are taken together—the complete case must be built.

back to top  Getting There Effectively

Under this view, the problem of new ideas languishing on shelves in research labs can be seen as going too far on one of these dimensions without considering the others. Thus—as we often hear at CHI—cool technology that is not usable or useful to the user is of no value. But similarly—something we hear less often at CHI—a usable, technologically-feasible product concept for which no business case can be made (or which is not strategically aligned with your company), is of no value, simply because it will never make it to market through your company. For example, the concept may be for a market that is too small to matter to your company, or the margins are not there, or it requires the company to become expert in something in which it will never be a market leader.

Companies have research departments so that they can explore possibilities at lower costs than would be possible with development divisions or in the marketplace. The goal for researchers who want to design products for real sale by real companies is to avoid getting too far away from the sweet spot in any direction.

bullet.gif Multi-Perspective Teams

Given that we are trying to get into the sweet spot on all four dimensions, we must arrange that our experiments address all four dimensions. Pitney Bowes is attempting this by creating collaborative teams composed of researchers and developers. The researchers usually include technologists (e.g., physicists, mathematicians, mechanical and software engineers) and user-centered innovators (e.g., anthropologists, designers, HCI specialists). But the team must also include those who can address the business and strategic value of the ideas (e.g., marketing, finance, manufacturing, sales, and operations). These teams must work to iteratively innovate, develop, and evaluate concepts against this full spread of needs.

bullet.gif Directions

At the same time, the teams must be careful to ensure that their evaluations are not premature or off-base. That is, one may miss the creation of disruptive technology by mis-applying current values (e.g., copier developers missed the fax machine because copy quality was so low) [1]. Managing how to move in this richer design space is a practice we are developing.

bullet.gif Tensions

Because the expertise in these areas comes from both research and development, there is a natural tension between innovating products aligned with the current way of doing business or in often uncomfortable new directions. Business runs the railroad; it is naturally interested in innovating for trains that are faster, better, cheaper. Research looks to make new markets; it is interested in innovation, in making tracks in a new direction. In order to stay near the sweet spot, we are finding it valuable to avoid trying to reduce this tension. In fact, we see it as a critical resource to addressing all the needs simultaneously. It helps get everybody moving out of their comfort zones: It keeps the researchers grounded, and the business folk stretching.

bullet.gif The Innovation Pipeline

As product concepts evolve, the difficulty of asking questions also grows. That is, we ask the simple questions first and if the idea still flies, move on to the harder questions. To support and encourage this, we have structured the work into a two-part "pipeline." A strategic question starts the work in the Concept Studio, where concepts are explored broadly around the question. If no concepts are discovered that appear to be in the sweet spot, the engagement is terminated. However, if likely prospects emerge, they are pushed to develop a crude value proposition and business case. (This can take from two to six months). The best of these are then passed to the Systems Laboratory for field trials. Here, the explorations are deep rather than wide, aimed at refining and proving out all dimensions of the concept: technological feasibility, user, business, and strategic value. At the end of this work (six months to a year), either the attempt fails or a business case can be made. Because the experimentation in the Systems Lab has addressed the risky issues, this resulting business case should be adequate for the partnering business unit to make a solidly grounded decision on whether and where to enter the concept in its development mix. As the concept moves through development in the business unit, research continues to consult, particularly in areas of significant novelty, to maintain the design in the identified sweet spot against the pressures of wanting to revert to the current comfort zone.

back to top  Design Cases

AC&T's experiment with this Innovation Pipeline started in the summer of 2001. Since then, roughly dozens of strategic questions have been explored, many concepts with appealing value propositions have been created and tested, and many opportunities have entered into development. These include a stamp-like postage dispenser that uses a novel security model; a system for protecting against anthrax-contaminated mail at a fraction of the cost of traditional contamination-proof workspaces; and an RFID-based practice for document management with significantly-reduced time for locating lost files.

We are learning firsthand how hard it is to keep everyone happy. But that is exactly why new-product development is tough. With our Innovation Pipeline, we believe that we are attending to the whole problem throughout the innovation process in a coherent and reasonably efficient way. By integrating the many needs that must be met to create a technically feasible product that is valuable to users and will make a good business for our company, we are sharply increasing Pitney Bowes chances of getting really useful products into real users hands. Central to this success is including business considerations in the framing of the HCI mission: not just innovations based on user needs, but good businesses built around such user-centered innovation.

back to top  Acknowledgements

My thanks go to all my colleagues at AC&T who are working so hard on this meta-experiment on product research. Specifically I would like to thank Jim Euchner, Brian Romansky, Cheryl Picoult-Triska, Gary Hansen, Yuling Wu, Trysh Wahlig and Pat Sachs with whom I have shared many discussions on all of this, and who are my partners in running this experiment.

back to top  References

1. Christensen, C. (2001). The Innovator's Dilemma. Cambridge, MA: Harvard Business School Press.

back to top  Author

Austin Henderson, Director, Systems Laboratory Advanced Concepts & Design Pitney Bowes


Kate Ehrlich
Collaboration User Experience Group
IBM Research
One Rogers Street, Cambridge, MA 02142

Austin Henderson, Director,
Systems Laboratory Advanced Concepts & Design Pitney Bowes
35 Waterview Drive MS 26-21, Shelton, CT 06484

back to top  Figures

F1Figure 1. The Product Sweet Spot

back to top 

©2005 ACM  1072-5220/05/0100  $5.00

Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.

The Digital Library is published by the Association for Computing Machinery. Copyright © 2005 ACM, Inc.

Post Comment

No Comments Found