Scoring Innovation

To start our list of initial public offerings, we looked to The Technology IPO Yearbook, 9th ed. (Morgan Stanley, New York, 2003), which lists, from 1980 onward, detailed information on each high-tech company making its initial public offering of stock. The yearbook gives the company's valuation at the time of its IPO, whether it was acquired or went bankrupt, and its market valuation as of 31 December 2002.

Of the 1303 high-tech IPOs listed there from 1993 to 2002, we first subtracted corporate spin-offs and other IPOs that weren't technology start-ups.

We divided the 823 remaining IPOs into a 1993 to 1996 baseline and a 1997 to 2002 follow-on period to measure innovation over time. This second period of six years also corresponds to an upsurge in VC spending, which in theory should have resulted in increased innovation.

To test for innovation, we rated each IPO on a scale of 1 (best) to 5, based on the following criteria:

No. 1: We reserved our top score for technologies representing a fundamental departure from anything existing previously and whose commercialization made possible an entirely new (and important) business market. Examples include bringing to the marketplace xerography, the microprocessor, the Web browser, public-key cryptography (VeriSign Inc.), distributed caching software for Web servers (Akamai Technologies Inc.), and high-temperature superconductors (Illinois Superconductor—now ISCO International Inc.).

No. 2: We gave this rating to companies able to demonstrate fundamental technology improvement in an existing product category. These include "disruptive technologies" that supplanted technologies in established markets. Ciena Corp. is a second-tier innovator, because it was one of the first to successfully manufacture wavelength-division optical multiplexing equipment to the exacting quality standards of wide-area network telecommunications customers, such as Sprint and MCI. Previously, optical telecom equipment could not handle multiplexing and so had many fewer communications channels.

No. 3: We gave this designation to companies able to demonstrate nontrivial technical improvements in existing product categories. These improvements generally extended existing technologies (for example, using ASICs with 0.13-nanometer instead of 0.18-nm traces), rather than deploying truly disruptive innovation. Juniper Networks Inc. is a typical tier 3 company—it developed packet-switch router hardware and software, which it then sold to telecom carrier customers, instead of doing a complete redesign of telecom carrier-class router hardware and software.

No. 4: Our fourth tier consisted of companies able to demonstrate modest improvement in existing technologies, perhaps by repackaging a combination of already commercial technologies. For example, Palm Inc., which went public in 2000, earned a 4, since the fundamentals of the personal digital assistant haven't changed much since the introduction of the highly innovative but commercially unsuccessful Apple Newton in 1993.

No. 5: Companies in this tier did not create new technology but were able to successfully market existing technology, or they developed new business models using well-established Internet technologies. HotJobs.com Inc. illustrates this theme: many other companies were using Web portals to post résumés and recruit individuals well before HotJobs went public.

You can find a complete list of the companies and their ratings at http://signallake.com/publications/IPORankings.pdf