Welcome to Web-resident business

The answers to these questions are important, as they dictate whether the spoils will accrue proportionately to the industry?s incumbents, or whether new companies and new business structures will emerge.

In factory automation, the first step is to optimize the production flow before introducing robotics. The problem with information-age automation is that the substrate is a shifting sand of innovation itself. Rather than a physical world with stable rules of physics, we are automating a virtual world in which the basic assumptions of connectivity, latency and time-space are subject to change.

Imagine how factory automation would change if suddenly, all machines could be interconnected with equal ease, products could move instantaneously from machine to machine, and copies of any product could be spawned at any time. Certainly, the factory design and layout would change.

Even more interestingly, what if trucks and trains could move a product cross-country as fast as a conveyor belt could move it 10 feet? Trading practices and the scope of vertical integration would change. Partnerships with separate companies would be feasible for all process steps in the production flow. Small companies that focus on a core of valuable services would proliferate. Adaptive businesses would be radically restructured.

This is not some fantasy, but a description of information-age business. With the Internet, information products can spread globally with near-zero latency.

Most modern businesses are information businesses at their cores. The optimal scope of an information technology business is defined by the connectivity and latency of its information systems. Business boundaries historically have been reinforced by the relative hassles of wide-area connectivity and proprietary network integration. Companies have become islands of automation.

An analogy can be drawn to the evolution of programming models, from desktop computing to client-server to network computing. This shift in program partitioning between desktop and server is a byproduct of the shifts in network connectivity and latency. Just as client-server topologies have become obsolete, so, too, with traditional business definition.

In the past five years, we have witnessed a radical change in the use of computers, from computation to communication. By improving real-time communications, computerization can improve most business processes, including sales and support, which are fundamentally communication exercises. Trade is a form of structured communication. Excess inventory is a byproduct of poor information flow.

Email was the Internet's first "killer app." It was the first ubiquitous communication channel and will continue to serve as the transport media for many business communications. Many businesses are just starting to tap into this rich channel of communication with their customers, using products from companies such as Digital Impact, Kana Communications, Tumbleweed Software and Tacit Knowledge Systems. In the next three years, we expect that 90 percent of direct customer interaction will be through email and the Web.

Businesses are starting to use the Internet to tie their business automation systems together. Twenty years ago, you could not assume that your trading partners in any business had computers on their desks. Ten years ago, you could assume that many of them did, but to hook those PCs together would be an expensive custom-development effort. Only recently can you count on a lingua franca that works throughout all businesses. My Macintosh can talk to your PC; we can start building business applications without having to worry about the plumbing. And that is a remarkably liberating substrate for business e-commerce.

Computer automation migrates progressively from the desktop to the enterprise to the industry. Once automation is widespread at one level, the opportunity to automate at a higher level of abstraction becomes possible. We have seen a similar pattern playing out in three of Draper Fisher Jurvetson's portfolio companies: Tradex, Magnifi and Saltare.com.

Tradex initially automated the purchase-order approval process. By replacing a paper-based system, it found it could reduce the overhead cost of a purchase by 33 times. Once procurement automation was adopted within a company, Tradex could offer connections to suppliers' catalog systems and enable intercompany trading. Once several companies were in the network, Tradex could promote industry-level automation by aggregating small-business demand, gaining pricing leverage with suppliers. (Tradex recently was acquired for $6 billion.)

In the field of marketing automation, Magnifi initially automated the enterprise by building systems that helped people search and retrieve rich-media files within a corporate intranet. But most marketing operations, from public relations to media buys, are outsourced to partners. The economic leverage of marketing automation is not inward-looking, on the desktop; it is in efficient partner communications. Magnifi found a much larger opportunity in automating the sharing of marketing materials across an extranet.

In supply-chain optimization, first-generation companies like i2 Technologies and Manugistics delivered systems for improving inventory levels within a company. Saltare.com has developed a real-time supply-chain optimization system that runs up and down the entire supply-chain extranet. By sharing information more widely between companies, Saltare's distributed intercompany application enables a much larger economic gain in efficiency and inventory reduction.

Once Saltare automates a large portion of an industry, a new level of efficiency and industry automation is possible. Industry capacity, timing and pricing can be rolled into a higher-level abstraction--a commodities market. Suddenly, a futures market for production goods can restructure the basis of production planning.

The economic leverage of computerization grows at each of these steps to a higher level of automation. There is much more to be gained by industry-wide trading networks than by further improvements in personal desktop productivity. Email is more important than Excel. Extranets are more important than enterprise resource planning (ERP) systems.

A new group of businesses, called application service providers, are hosting these enterprise and industry-level applications. Entrepreneurs are not starting "software companies" any more. They are developing Web-based services accessed through a browser. These hosted applications allow for frequent (even daily) invisible upgrades, access to real-time data feeds, and easy release across all corporate desktops and remote "Webtops"--at home and overseas.

Much of the early leverage for these Web services will be with small businesses. Small businesses will dominate the U.S. economy, yet today they are islands unto themselves. Companies such as DigitalWork (Web services aggregator) and Everdream (subscription computing for small businesses) are finally forging effective channels to reach small businesses over the Web. With channels in place, small businesses can gain from the sophisticated services that used to require a direct sales force and a complex corporate installation. Procurement automation, marketing automation and sales-force automation will finally be available to small businesses.

The business landscape will undergo further tumult with auctions, dynamic pricing, multilingual trading exchanges, outsourced e-commerce, and a variety of innovations that were not possible in the physical world. It will not be business as usual. And as always, with radical change, new companies will be the first to exploit the new paradigm. Like the consumer Internet powerhouses of today, most of the business e-commerce winners will not be companies that existed five years ago.