Phil White reveals the Informix recipe for Internet success.
On December 20, 1995, Informix Software Inc. chief executive Phillip E. White announced that his company would acquire Oakland, California-based Illustra Information Technologies. Just six weeks later, Informix announced its product strategy for the Informix Universal Server, a merging of Illustra's extensible data type ("DataBlade") technology with Informix's core parallel database technology. The goal of this merger is to deliver a DBMS that can manage all of a company's information assets, including traditional structured alphanumeric data, still images, sounds, video, long text, and other kinds of complex data, including Web pages.
In phase one (during the second quarter of 1996), Informix will provide a gateway that lets customers build applications that integrate complex data stored in an Illustra Server with traditional data stored in an Informix DBMS. In phase two (also during the second quarter of 1996), Informix will deliver a DataBlade Developer's Tool Kit for creating new user-defined data types that run against both Illustra and the Universal Server. In phase three (during the fourth quarter of 1996), Informix will deliver the Universal Server, consisting of the fully merged technology, complete with "snap-in" DataBlade modules.
While the implications of the Informix Universal Server extend far beyond the Internet, it is the Internet opportunity that drives the Informix strategy. White believes that the explosion in user population and the proliferation of low-cost, global networking will accelerate the demand for data. He plans to satisfy that demand with technology that can deliver any kind of data -- simple or complex -- to any user, anywhere in the world.
Internet Systems Publisher David Kalman recently interviewed Phil White at Informix's Menlo Park, California headquarters. In the course of the discussion, White offered his thoughts on Internet technology, database technology, and his company's role in the next computing era. For added effect, note that this interview occurred the day before Informix announced its 1995 results. Fourth quarter 1995 total revenue increased 45 percent (to $217,070,000) and net income increased 63 percent (to $38,834,000) over fourth quarter 1994. For fiscal 1995, total revenue increased 51 percent (to $708,985,000) and net income increased 59 percent (to $105,333,000) over fiscal 1994. In the context of this stunning financial performance, you ignore at your peril Phil White and his Internet-enabled database company.
KALMAN: Today, when people think of the "Internet," they think of Web surfing and static information publishing. If Internet technology is to be adopted by IS for corporate applications, how will we get to that point?
WHITE: It will happen just as the PC phenomenon happened. PCs started out with individual users at work and at home; now the PC is the client of choice. The Internet will do the same thing. With lots of users, easy access, and low cost, IS has the ability to deploy applications to a larger number of users, quicker, faster, and easier.
It's happening in the right succession. First, you want to get everybody on the Net. You want everybody to understand the Net. Then IS can come in and satisfy user demands with Internet-enabled applications such as those for accounting, expense reports, HR, and customer service. IS now has a great opportunity to do a lot for the end-user community.
It's going to take lots of work to provide performance, ensure security, and support new media types. It's a great opportunity for the IS organization to step up and for the first time really increase the productivity of all the users inside a company.
How does Internet technology address the need to reduce computing costs in corporations?
I don't think it's a way to lower your IT expense. Rather, it's a better way to utilize your IT expense, and address more opportunities.
Instead of building large applications, companies are now buying large applications. Instead of building new mainframe applications, they're deploying new client/server applications. The intent is to make end users more productive. The Internet is a great way to do that at lower cost. That's why you'll see a phenomenal number of new server-based applications. The Web enables people to get to data more quickly. On the server-side of the business, we'll Web-enable everything we do and deliver data to Web end users.
I've read some headlines that call Internet technology the death of client/server. ...
There will always be a phenomenal number of client/server applications. The accounting organization will have accounting applications. The customer-service organization will have customer-service applications. But, all of those back-office applications will be Web-enabled, so casual users can get to accounting information and customer support information. The mainstream users of those traditional applications will continue to exist. I don't think [Internet technology] infringes on client/server. It widens the opportunity for more people to get at the data that currently exists in client/server environments.
Over time will today's traditional client be another kind of client?
Probably. You could have a mobile client or a desktop client. A client could come across the Internet or via another communications mechanism. The Internet offers a lower cost of deployment than client/server, and that's good news because those lower-cost users want more applications and more kinds of data. That's why the Internet is good for the server-based business.
The investment community and the press have focused on Internet technology startups. Do these companies have the technology and expertise suitable for enterprise-class applications?
Internet technology will appear in waves. Most of the people in today's client/server environment will also be in the Web environment. Many new startups will not make it. It will be difficult to figure out how to make money in the Internet environment. Extensions of today's technology will be one way to make money on the Web. Many of these startups will be acquired by bigger users; they will merge with other kinds of technologies. A few will make it, but I think it will take a couple of years to shake out. Today, it's still difficult to figure out how to make money -- enough money to sustain long-term growth -- on the Internet. Even the biggest players, such as Spyglass and Netscape, aren't making enough money to maintain long-term viability, although their stock prices are high. Over the next couple of years all that will sort out and people will figure out how to make money and have a sustained business.
Do you see a real incentive for organizations to rebuild applications using the Internet architecture?
I don't know if you have to rebuild applications. You certainly want to extend your applications to provide Internet access. Java and Java-based development tools are coming on fast. Everyone's building Java derivatives. There's LiveWire and LiveWire Pro from Netscape. There's Latte from Borland. Everybody's got a derivative. Everybody's trying to get in. It may be the next paradigm in programming, but to succeed it will require larger companies with more staying playing power. Java will be successful because Sun is behind it. There are many companies picking up Java to build products. Are the browsers the key? No. The browsers are the first entree to the Internet. Browsers will be totally different in the next couple of years.
Internet-style applications are in some ways like mainframe applications. Is this technology something new, or are we just being reintroduced to something we knew all along?
What's new is the access. When you get a home page, it's static data with color and graphics. The real opportunity will be to add animation and the kinds of things that make those applications more productive. The real opportunity is the low cost of the Internet, versus the cost of other communications. You can get more people on the Internet than you can on LANs and WANs. In that sense, it will drive more and more of our kinds of business in the server database side. Users will look for more kinds of data and more kinds of applications. That's one reason why we acquired Illustra.
I wrote a DBMS column in which I called Informix "the last true database company." Is this still true?
Our core businesses are still application development tools and databases. The difference is that we've found new and creative ways to extend those technologies. Our database is now extended to provide new data types for the Web, for SmartCards, for data warehouses, and for multimedia. And, our development tools support these new extensions. So now our application development tool can do SmartCard applications, mobile applications, wireless applications, and data warehouse applications. We're still data-centric. We're not in the consulting business; not in the communication business; not in the application business. So we still are the only true database company. Our core technologies are designed around databases.
The difference between us and our competitors is that we have a product line. All of our competitors have lines of products. Their lines of products get them into new markets. Our goal is to extend our product line into new types of opportunities without destroying our core business: database technology.
What is the opportunity for database technology? Many people view DBMS technology as a commodity item.
Over time, the database will be the most important piece of IT. It's what stores, maintains, and allows access to all the data on which a company runs. Today we manage less than 20 percent of the data that a company uses day to day. With a relational-object environment, we can handle 100 percent of the data.
Imagine if you had all the data that's stored in emails, voice mails, faxes, filing cabinets, and current databases, and made it accessible inside companies using a low-cost environment such as the Internet. The database then becomes the most important piece of technology that the company will acquire. That's why we can continue to extend the database to handle new data types, stay focused, be the only pure database company, and continue to grow our database business at 50 to 60 percent. And our competitors' database business slows down because they're focusing on many other things. It doesn't mean their business model is bad, it just gets back to your point.
We're still the only true database company in the industry, with no new companies coming into the market. And, the database business is growing faster than any other segment and is more important than any other segment. That importance takes on a whole new light when you can make people more productive by giving them access to data stored in a number of different areas.
You recently announced your Informix Universal Server. What is the origin of this technology?
The Universal Server is a merging of the architectures of Illustra and Informix. We were extending our [Informix] database to handle new data types using SQL3. That work allowed us to integrate the Illustra functionality much more quickly than anybody believed we could. We think the combination of the two -- in a product we call the Universal Server -- will give us an 18- to 24-month lead in the market. A year-and-a-half to two-year lead in this marketplace is a lifetime. The term "universal" refers to the data types. This server handles all types of data -- text, image, voice, graphics, and more. We believe that the Informix Universal Server will be the database of choice for the Internet.
By "universal," are you also implying that one DBMS will fit all applications, including OLTP, decision support, and so on?
Over time, we'll look at merging everything into one product. We'll have a massively parallel product, a clustered SMP product, and a workgroup product, but all with a common architecture. The real issue is not if we can do it. We can do it. The real issue is the performance. Can we drive thousands of transactions per second with the Universal Server? We think that's as much a hardware issue. The price of hardware is coming down dramatically, the performance is going up dramatically.
The data is also growing dramatically.
That becomes an issue of how you compress and decompress these big pieces of data such as images. Those aren't OLTP kinds of data. Usually OLTP kinds of data consist of numbers and characters. We think we can get that as fast on the Universal Server as on today's relational product. I think over time we'll have one engine that fits all, and it will be as fast as the standard relational product, with the benefit of handling all these new data types without reprogramming.
We believe that the real opportunity for a technology company like ours is not to do things faster, but to increase the productivity of the users. I don't think many companies today offer that benefit. Everybody talks about peak speeds, more graphical, and more "client/server," but the ultimate measure is the users' productivity.
Productivity is what everybody wants. Everybody's afraid to talk about it. We're going to talk about it. We're talking about the traditional kinds of applications that are client/server- or even mainframe-based, and allowing them to be more productive and deliver more value without a total rewrite. We'll increase the users' productivity, taking advantage of new things such as the Internet, with minimal disruption.
Can you explain the SmartCard technology you mentioned earlier? What is it?
SmartCard is a technology that we've announced jointly with Hewlett-Packard and GemPlus Card International Corp. It's a programmable, intelligent, secure card technology. You'll be able to track what you buy, where you go, and what you spend. We think it will be the next client of choice.
HP makes the card readers. GemPlus is the world's largest manufacturer of SmartCards. The [credit card-size] SmartCard contains a chip that can hold 8000 characters. The traditional magnetic stripe has 28 characters. Every 12 to 18 months, that capacity will double. In the next two to four years, we'll see 64K to 128K on the card. This card will fit in a PCMCIA slot and it will slide across a card reader on a keyboard.
This card will be used for buying, identifying, and securing things. We jointly developed the security with HP, and it has been approved by the National Security Agency. We have security by PIN, and we'll have biometric security. You'll have personal security because you'll put your finger on a space on the card, and it will validate you off the chip. Security is a big issue on the Internet, but once we have security of data -- more than just a PIN number -- then electronic commerce will flourish across the Internet.
Can businesses rely on a public utility such as the Internet to run mission-critical applications?
That's where phone companies will play a big role. Every phone company in the world is Internet-enabling their business applications and technology. The phone system will ensure that the communications issues are addressed. The cost of the communications will be so low that you can retry and recover quickly.
What's your sense of the size of the market for Internet technology?
The Internet has caused a bigger explosion on Wall Street in stock prices than it has in revenue generated by companies involved in it. Netscape and Spyglass have small revenues -- probably negative revenue flows -- with phenomenal market caps, in the anticipation that they'll find ways to make money and dominate their market segments. That will take a couple of years to see.
They're certainly opening up opportunities for us all. More users make more business for all of us. Can you make money off it? Yeah. Long term. In the short term, it's a phenomenon because of the way it got started. But, nobody pays much for a browser. Nobody pays much for access. And the applications are pretty static.
The other way people make money is to charge for advertising, not for accessing data. Going forward, that will all have to change. There's a whole other wave of people who want to access more kinds of data. Today we make money on a per-user basis. It's difficult to charge per-user when you don't really know who your users are. We have to figure out new paradigms of how to make money, and how to charge for these services.
What do you think Microsoft's role will be in this market?
Microsoft will be a worldwide player. They were late to the market, which is a good indication that Microsoft may not be the most strategic company in the world to have missed something like this. But they execute pretty well.
I could argue that acquiring Illustra now makes Informix relatively late to market with Internet technology.
I'd say it the other way around. The Internet gave us the opportunity to take advantage of what Illustra has, and get ahead of the market.
The Internet today is static. The opportunity is to add the productivity and value inherent in animating the Internet. Everybody's rushing to build browsers and tools, and forgetting about all the data that these tools will need. The opportunity is right there ahead of us. That's why we have a good one-year to two-year lead on everybody else. Everybody else is worried about the browsers and Java. Not about the data.
The Internet opportunity will be to provide servers for all the new kinds of data, replacing voice mail systems, email systems, filing cabinets, and so on. We were building this stuff, but Illustra allowed us to do it faster by merging our two technologies. More important, acquiring Illustra gives us the key to our future: the dream team of development. We now have two world-class development scientists. Mike Saranga built the best relational database in the world -- DB2 (and now he's doing it for us). Michael Stonebraker built a great technology company in Ingres, and then built Illustra. Microsoft doesn't have them. Sybase doesn't have them. Oracle doesn't have them. IBM doesn't have them. But we do.
We have the smartest minds in the database business. We can innovate faster. We can do it quicker. We can do it smarter. That's what will keep us ahead. That's what will keep us as the fastest growing database company. I don't have to go hire anyone to figure out what someone else is doing. I don't worry about what Microsoft, or Sybase, or Oracle does. I've got people who've forgotten more about databases than most of our competitors know.
Informix Software Inc., 4100 Bohannon Dr., Menlo Park, CA 94025; 415-926-6400; http://www.informix.com.