Deconstructing the Web

Baldwin Louie — Mon, 14 Aug 2000 15:00:00 -0700

Once you figure out your e-business strategy, it's time to start designing the Web site, right? Not any more. The era of the browser, in which the world's information is linked together through Web sites, navigated with search engines and collected and reorganized through portals, is coming to an end. Over time, it will become an increasingly smaller part of what the Net economy offers those seeking competitive advantage.

If you focus your efforts entirely on today's Web-based architecture, you'll cut yourself off from an explosion of new devices, interfaces and software technologies that are proliferating at an accelerating pace.

Even if your customers, suppliers and other stakeholders are thrilled today to interact with you exclusively through the Web, it won't be long before they'll demand new services and new interfaces that operate through PDAs, cell phones, pagers, instant messenger clients and other computing devices. And not long after that, it will be time to communicate directly with noncomputing devices, such as cars, refrigerators and packages.

Indeed, the current investment mania for all things wireless and for scalable Internet infrastructure is built on a belief that the proliferation of connected devices will spawn new generations of killer apps. I share that belief.

The disruptive agents here include our old friend Moore's Law, which continues to drive the creation of smaller, cheaper and ultimately disposable client devices. At the same time, common data exchange standards like XML continue to spread and evolve. The result: an explosion of new opportunities for communicating information between everything and everything else. Few things in life (and in business) are as predictable as these trends.

The Web is disintegrating into bits. When the dust settles, what's important for a successful long-term strategy is not Web site design but the flexibility of your information architecture.

SCHIZOPHRENIC USERS
As we enter the next generation of information architecture, some of the Internet's pioneering companies, such as America Online (dossier), will find themselves saddled with their current Web-based interfaces. They won't work. A television is neither a radio with pictures (as it was first described) nor a diskless computer (as computer makers describe it today). Just as a car is not a horseless carriage, a car is also not a browser with wheels.

New Internet devices and applications are different not only because they are designed to do different things, but also because the mindset of the person who uses them influences the kind of interaction that makes sense. You cannot shove a Web page onto a cell phone's tiny screen. Even if you could, it wouldn't make sense from the user's point of view. The information I want delivered to my cell phone is different from what I want on my home computer. Web pages are designed for sedentary, reflective use. The phone is for urgent, short communications.

As Microsoft (MSFT)'s Rick Beluzzo put it at The Standard's recent Internet Summit, we are entering an era of "technology schizophrenia," where the same person takes on a different profile and personality with every device and application he uses. Each user takes on a role appropriate to the device, and the roles have different requirements for depth of information, timeliness, interactivity and speed. You will need a different information interface for each role surfers play.

THE AGENT ECONOMY
I recently became senior strategist for a startup company called Spyonit that is developing technology to serve "schizophrenic" users. Spyonit turns the browser model on its head by focusing not on Web design but on the underlying information and how it might be used. The company's technology lets its customers specify information that is important to them, what changes in this information they want to be told about, and how and where they want the information to reach them.

As a consumer, for example, I can create "spies" to watch the various auction sites and let me know when items I want are for sale, or when the status of a particular auction changes. The spy can notify me by e-mail, a custom Web page, PDA, cell phone, instant messenger or pager. For each device or client app, the notification is tailored to the device and my role. If I ask to be notified by cell phone, the spy knows that this is an interruption, and gives me only a headline. ("You've been outbid on Richard Armour's Twisted Tales From Shakespeare.")

Spyonit customers can also use the technology to deepen relationships with their business partners. Morningstar (dossier), the online financial information giant, has embedded spies on its site that lets customers signal how and when they want to be notified of changes to their portfolio. At the end of the day, it can send me an e-mail with the closing prices of all the funds I own; but if, for example, my technology holdings ever go over 30 percent of my total portfolio, I can have Morningstar page me. Each alert provides the company with opportunities to interact with its customers.

This is the first step toward an information architecture in which the "browsing" isn't done by individuals, but by software agents that are delegated to watch a variety of conditions and report any changes. From here, the natural extension is to charge the agents with certain actions beyond notification. For example, in the future I might create a spy that looks to see if Patrick McGoohan's The Prisoner is going to be on TV. If it is, it can program my VCR (or my computer, for that matter) to record it.

THE IMPORTANCE OF PLATFORM
In the deconstructed Web, what matters most is not the site design, but the information design. The more your data is structured in abstract forms, stored in standardized formats and built on hardware and software platforms that can scale to large numbers of users and massive amounts of information exchange, the better your chances for a sustainable e-business strategy.

In part, this is because no one really knows what the future applications of Internet technology will be. Often, even application inventors are surprised by the unintended uses that evolve (sometimes quickly) for their creations. It is human nature to start out believing new devices are merely improvements over existing ones (again, the horseless carriage and radio with pictures). With experimentation and serendipity, new applications emerge. The less you embed the old metaphor in your strategies and systems, the easier it will be to respond to new opportunities.

PlanetRx, for example, has given some of its customers handheld bar-code readers from Symbol Technologies (SBL) as an experiment in interactions. Customers scan the bar codes on the things they're running out of to create their shopping list, then zap the list to PlanetRx. PlanetRx fills the order and sends it to the customer.

Consumers should welcome this new way of ordering because it doesn't change the interface they are familiar with; it only reduces a bunch of annoying steps (few people find going to the drugstore fun). Don't write down your shopping list, just point to what you want. Anybody can understand that.

Aside from turning the advantages of physical retailers (location, location, location) on their head, this application has the added power of starting the process of collecting new data in bulk and in standard form. If the platform is properly constructed, PlanetRx can discover and implement new applications as they present themselves.

Maybe the data will flow back up the supply chain, improving product forecasting (from weeks to minutes) or product development (what's not getting used or what is bought with what) or even the ordering of raw materials (make more plastic, make less plastic, print more labels). Eventually the products will just reorder themselves.

The key is to build around the platform itself rather than the devices and their associated roles.

CATCHING THE WAVE
As you think about your e-business strategy (or its next generation), here are a few ways to prevent value from being stuck inside your Web site:

Understand the different roles users play with different devices and applications, and develop interactions that are appropriate to each one.
Design information separately from the interface. In fact, start with the overall information architecture, not the look and feel of the first-generation application.
Build the applications themselves on a hardware and software platform that can scale. Assume that uses that have not yet occurred to you will become strategic sources of competitive advantage.

Larry Downes is an e-business consultant and coauthor of Unleashing the Killer App: Digital Technologies for Market Dominance (Harvard Business School Press).

The Industry Standard - Deconstructing the Web - Comments

Deconstructing the Web