David W. Lewis

Sustaining technologies improve the performance of established products along dimensions of performance that mainstream customers in major markets have historically valued. Sustaining technologies improve products or processes, and they can be driven by new, and sometimes even revolutionary, technologies, but what is important is that the improvements result in accomplishing the same thing, only doing it better. Because of this, established customers recognize the value of the improvements. Relationships in the market, cost structures, and organizational dynamics can thus remain largely unchanged.
Disruptive technologies bring a very different value proposition to the market. Generally, disruptive technologies initially underperform established products in mainstream markets. This makes them easy to ignore. But disruptive technologies have other features that are valued by a few fringe or new users. They also improve at a faster rate than established technologies. This is what makes them dangerous to established firms. Disruptive technologies often appear to be merely toys, but before you know it the toys have grown up and are cheaper, faster, and better than what established firms are selling.

7 thoughts on “David W. Lewis

  1. s.A

    >Established organizations are generally good at change that involves sustaining technologies. They know the needs of their customers and how to work with and listen to them. Service models are effective because they have been refined over long periods. Sustaining technologies lead to improved quality that makes it possible to justify higher costs.

    Established organizations generally fail when change involves disruptive technologies, and organizations at the periphery or from different sectors most often succeed. Products or services are designed for new, rather than established, users. Disruptive technologies are often developed before applications are known, and it is only after several fits and starts that applications for the technology and users of it are found. Invariably cheaper and faster, disruptive technologies are often easier to use even if quality is not high and capacity is not large at the outset. In general, disruptive technologies require new service models and pricing structures that challenge established organizations and the interests and expertise of the individuals within them.

    Reply
  2. s.A

    >COLLECTIONS

    Collections have been the heart of libraries for as long as there have been libraries. Many of the electronic products offered to and through libraries are sustaining rather than disruptive technologies. For example, electronic journals, especially when offered by established publishers or in collections such as JSTOR, don't fundamentally change the service model. Libraries still purchase output compiled by publishers and make it available to the libraries' users. Access and availability are enhanced, and the costs are a bit more, but the enhancements in service justify the increase. However, disruptive technologies are coming quickly. New standards like the Open Archives Initiative's Protocol for Metadata Harvesting make possible a new distributed model for the storage and retrieval of documents. This model will deconstruct the journal as it is currently known into individual articles in much the same way that music file sharing has deconstructed the record album. Repositories of e-prints, images, and other documents will use the developing standards to make these items available to the world. While peer review, citation analysis, and a clear understanding of the pecking order (which bestows the status required for promotion and tenure decisions) have yet to emerge, be assured that they will emerge. Like most disruptive technologies, repositories are significantly cheaper than existing products. Several years ago Odlyzko analyzed the cost of Ginsparg's physics archive and estimated that the cost of placing an article in the archive was $75 versus the $2,000 to $4,000 it cost to publish it in an academic journal. Odlyzko goes on to say, "As Andy Grove of Intel points out, any time anything important changes in a business by a factor of ten, it is necessary to rethink the whole enterprise. Ginsparg's server lowers costs by about two orders of magnitude, not just one." Many repositories also provide more access than library collections. They are generally available to anyone with an Internet connection at any time from any place. The power of this disruptive technology is seen dramatically in the story of Lubos Motl, an undergraduate physics student at Charles University in Prague, as recounted in the New York Times:

    Mr. Motl posted a research paper on the archive and the results were flabbergasting. Established string theorists were so impressed by his work that he ended up with a scholarship to Rutgers, where he is completing his doctorate. "I was at first a little annoyed by the first paper, because it scooped me," said Dr. Thomas Banks, a physicist at Rutgers and the University of California at Santa Cruz and one of string theory's prime innovators, who had been working out a similar idea. "This feeling turned to awe when I realized that Lubos was still an undergraduate." Mr. Motl is a striking example of how the archive is changing physics.

    Note that the physics archive is changing the discipline. Note also that libraries are not part of the story.

    Reply
  3. s.A

    >BIBLIOGRAPHIC CONTROL

    Librarians have great expertise and a long history of developing and using bibliographic control, but the systems in use were developed for technologies based on paper and have been migrated to an electronic environment through a series of sustaining technologies. There is still an assumption that an institution's catalog is a core-required activity. But in an environment where collections are housed in distributed repositories and harvesting services search them, what purpose does a centralized catalog serve? Would individuals not be better off with their own catalog of the items that interest them and updating tools that learn from the actions an individual makes while using this personalized tool?

    Change has also come from another direction. For the past twenty-five years, librarians have largely ignored research on algorithmic retrieval. The technology did not scale, and authority control and Boolean searching worked as reasonably effective on large files. And then there was Google. Ask any reference librarian where they start when answering a question–Google or the library's OPAC. Algorithmic retrieval now works and scales, and it is no longer reasonable to suggest that the Web needs to be cataloged using established library practice.

    Even in the more limited arena of books, there is the potential for disruptive change to overwhelm established library approaches to bibliographic control. Amazon.com is already used by many as a substitute for the library catalog. What will happen now that the full-text of many of the books in Amazon.com's inventory is searchable?

    Reply
  4. s.A

    >REFERENCE

    Since the early years of the twentieth century, libraries have provided reference services. Users with questions came to the library and got assistance from well-read generalists, or, in more specialized libraries, experts in a field. Upon entering the library, users were met by a friendly, helpful, reassuring librarian who could almost always find enough information to satisfy the user's need. This help was required in paper libraries because of their complexity. Even in the current immature Web environment, across the country the number of reference questions being asked in libraries is in steep decline. Even with email and chat services, fewer users seem to need a librarian's help. Google and other Web tools are easier and good enough, at least, for many questions.

    Even expert help from librarians might be replaced. Google Answers, which creates an e-Bay-like market place for people with questions and people who are prepared to answer them, is one alternative. A recent study by Kenney et al. at Cornell University indicates that Google Answers is significantly cheaper than comparable reference services at Cornell. For the easiest questions, Cornell was five times more expensive than Google Answers, and for the most difficult questions, Cornell was twice as expensive. The Google Answers responses, even when graded by reference librarians, were nearly as good as the responses of reference librarians themselves. The clear implication of this study is that Google Answers or similar services that harness the expertise of a large number of people in a question marketplace could easily be competitive with library reference services.

    Reply
  5. s.A

    >What is clear in the review of all three areas of library practice is that alternatives exist, or could easily exist, that are cheaper, easier, faster, and more convenient than the comparable services now offered by libraries.

    The past success of libraries, and the generally high regard in which they are held, will not protect them from potential competitors armed with new technologies. Past success with sustaining change should not provide libraries with confidence. Libraries are surrounded on all sides by disruptive technologies. Because of this there is much to learn from Christensen.

    Reply
  6. s.A

    >David W. Lewis is Dean of the Indiana University-Purdue University Indianapolis University Library.

    Reply

Leave a Reply

Your email address will not be published. Required fields are marked *