A Chat with the Master of P2P


He created the popular Lotus Notes collaboration program, the first big success in information sharing beyond e-mail and simple file transfer. Now, Ray Ozzie wants to build on that start. In October, 2000, he launched Groove Networks, a company that sells a peer-to-peer product that lets small groups share files, mark up a virtual whiteboard, or have an instant-message conversation -- all without running into problems from corporate firewalls. Ozzie, who is considered a P2P visionary, recently participated in an e-mail interview with BusinessWeek Online's Olga Kharif and Alex Salkever. Here are edited excerpts of that exchange:

Q: How did you get interested in P2P?

A:The epiphany of sorts was that I kept watching my daughter, Jill, doing her homework with her friends over [AOL Instant Messenger], and my son Neil playing Quake [a search-and-destroy game that can be run on a multiplayer network] with his online friends.

In watching Neil in particular, I found that Quake was an immensely effective collaborative environment for a shared task: His team had to "capture the flag" of the other team. It used every bit of horsepower of the PC and network to help each player be efficient and effective at that one task. In business, we commonly have projects that require multiple people to self-organize and solve problems. But why are we stuck using e-mail, when technology is being used to serve these kids so much more effectively?

Over the course of the '90s, I had the opportunity to visit many companies, and talked to many CIOs (chief information officers) and line-of-business managers who used Notes as well as Web-based solutions within their enterprises. During that period, I witnessed many businesses transform from vertically integrated command-and-control hierarchies into adaptive organizations that were trying to effectively integrate outside business partners and customers into their core business processes and practices.

But the communications and computing systems that we, as an industry, have built for enterprises reflect the "old" style of managing business: Inside vs. outside, us vs. them. Firewalls keep out "outsiders" -- including partners and customers. Critical and valuable applications are available only to "insiders," except when carefully exposed through narrow "portals." That isn't the way that people or successful enterprises work today! Effective organizations integrate partners and customers to such a degree that firewalls have become a barrier to cooperative work.

This reality pushes much of day-to-day, person-to-person communications outside of strategic and valuable "systems" and instead into e-mail, phone, and fax. I knew from Notes that interpersonal interaction could be far more effective and valuable to business if it were done in the context of an application environment tailored to specific interaction needs -- far more.

Q: How did that translate into Groove Networks?

A: My goal in creating Groove, simply stated, was to create a platform for direct, dynamic, private person-to-person interaction -- a platform that could bring corporations the strategic value of "systems" tailored to their needs, but in a way that is a natural, great experience for the people who use it.

When choosing technology for bringing people together fast, flexibly, and effectively, I again looked to Quake. One of the kids had been running the Quake server on his local PC, and the teams had self-organized. When we started to work on the Groove project in late '97, we went directly to "peer" technology specifically because it was the best way to solve the problem of getting self-organized groups together in a fast and secure fashion.

Q: What are the most promising applications of P2P? Why are they so promising?

A: Many industry watchers have categorized peer computing in three ways: distributed content management, distributed computing cycles, and person-to-person collaboration. I'm obviously biased, but I believe the area of most interest is collaboration.

That's where intellectual capital can be maximized. For example, GlaxoSmithKline works closely with outside research organizations -- university laboratories, clinical research organizations, and biotech firms. The interactions between GSK and these external parties are extremely confidential and include complex conversations, document sharing, and image sharing. Shrinking just a few days off the discovery or trial period for a new compound can have enormous bottom-line implications.

Q: Why is enhanced collaboration important? Aren't the existing levels sufficient?

A: Companies are always looking to attain competitive advantage. These days, competitiveness increasingly depends on how effectively and efficiently a company interacts with its partners, suppliers, and customers.

Companies increasingly are focusing on their core competencies and outsourcing other elements of the business, whether it's manufacturing, logistics, or distribution. So technologies like Groove that address how people in different companies, or within the same company but in different divisions, connect with one another quickly and securely to get real work done is attractive. The goal of these companies is to reduce their "cost of connection." That's the main point. Peer computing can address these needs more effectively than existing solutions.

Q: What are the areas that won't work for P2P? Where is it ill suited?

A: I would not use a peer technology for a corporate purchasing system, a knowledge or enterprise document repository, or for an airline reservation system. I would, however, use peer-based systems to add a dynamic "interaction layer" surrounding those systems to assist in the purchase-approval process, the collaborative process surrounding the creation of documents, or in the communications among travel agents and consumers.

The real answer is to leverage centralized and peer-based systems for their strengths, and integrate them where appropriate. Just as the PC, whose 20th anniversary we are celebrating this year, didn't supplant the need for mainframes, neither do peer systems supplant the need for centralized systems.

Q: What are the best ways of making money from P2P? Has anyone come up with a winning business model yet for either the consumer or enterprise space?

A: I don't mean to be flippant, but the best way for any technology company to make money, whether it is based on peer-to-peer technology or not, is to deliver true economic value to its customers. If you provide true value, then customers will pay you for that value. Then our job is to ensure there's a healthy gap between the price we charge for that value and the cost we incur in delivering it.

Q: Why has P2P largely remained confined, in the consumer space, to instant messaging and music files? What needs to happen to make the bigger jump to broader applications?

A: Confined isn't necessarily the word I would use to describe the use of peer-to-peer technologies among consumers since instant messaging is growing exponentially, both among consumers and within businesses.

The initial opportunity may be to provide additional value beyond the instant-message platform. For example, earlier this year we participated in Microsoft's Hailstorm announcement. There, we demonstrated how MSN Messenger users could click on a Groove icon from within Messenger to open a Groove shared space.

In that space, users could work together more effectively on a high-school history project, or in a business context, work together on a market segmentation plan for a new product. In this example, MSN Messenger users are able to take advantage of our local storage, our support for on- and off-line use, and our end-to-end security, to name a few.

Q: What's missing from current P2P technologies (including your own)?

A: That's an awfully broad question. The lack of standards might be the obvious answer. For example, a standard that allowed all instant messaging systems to communicate with one another, just as e-mail systems do, would further accelerate adoption and increase value.

Another way of looking at it is: How rapidly will peer technologies be integrated into existing business infrastructure? For example, the PC really increased in value when it became integrated with existing computing systems. The same can be said about the Web. Browsers and Web servers were popular in the early days of "brochureware," but acceptance and growth accelerated dramatically when they became woven into the fabric of existing business systems. Peer is not an exception. On its own, it already has achieved some success. But when you see customers and existing system vendors making efforts to integrate Groove or other peer technologies, you'll know that an important threshold has been crossed.

Q: When will corporate P2P collaboration take off? What will drive it?

A: It's already taking off. A recent Media Metrix report indicated the number of at-work users of IM has increased by 22% in the past six months. Business imperatives to increase efficiency, reduce new-product time to market, and service customers more effectively and efficiently will drive adoption of peer-to-peer collaboration technologies such as Groove. As I mentioned earlier, we are operating in an increasingly decentralized business environment that demands more decentralized collaboration technologies.

Q: What is the ultimate, supercool future of P2P, in your own wildest dreams?

A: Rather than address peer-to-peer in general, let me talk about Groove more specifically. One thing that gives me the greatest satisfaction is when people I've never met -- and might never meet -- are able to take what we've developed and create a business around it. They take the product in directions none of us could have. That's incredibly rewarding, the way it brings people together. At the risk of sounding too much like John Lennon, the ultimate, supercool future of Groove and other collaboration technologies would be to play a role, however small, in bringing people together.


Hollywood Goes YouTube
LIMITED-TIME OFFER SUBSCRIBE NOW

(enter your email)
(enter up to 5 email addresses, separated by commas)

Max 250 characters

 
blog comments powered by Disqus