Interview: Portal Software's John Little

Starting an online services company and calling it "Portal" may not seem particularly brilliant -- unless, of course, you did it back in 1985. That's when John Little, a Princeton University graduate and Bell Labs alumnus, founded the company then called Portal Communications. Now known as Portal Software, the Silicon Valley firm supplies the "infrastructure" for Internet business -- customer management and billing software.

Still riding high on the success of his IPO last May, Little was recently in Hong Kong, where he spoke at length with Computerworld Hong Kong Editor Don Tennant.

COMPUTERWORLD HONG KONG: I understand that Portal Software dates back to 1985, when it was known as Portal Communications, and that you were actually the first company in the world that was allowed to sell commercial access to the Internet. How did you manage to get such an early start?

JOHN LITTLE: A couple things happened in college that are really the genesis of this whole thing. One is I had the demonstration of what was called the Arpanet back then from a 13-year-old kid, and it was kind of a personal epiphany, because I saw that a person sitting in one place could use computers scattered across an arbitrarily large area. I'd seen the opposite -- one computer with time-sharing terminals. But I'd never seen one terminal with a bunch of computers.

I started to extrapolate that out, and said OK, eventually every single computer on the entire planet is going to be connected to the Internet -- that was 1976. Of course there was a lot of fuzz in this about how that would happen.

The other thing that happened is I read a book by a guy named Ted Nelson entitled Computer Lib and Dream Machines. Ted invented the concept of hypertext and so forth, and he was really pushing the applications of computers, not so much the technology -- particularly the concept of the universal hypertext document, where all the world's information is at your fingertips. Great concept, but how do you make any money off of that?

By the time I got out of college and was working at Bell Labs, what had become clear is that the infrastructure is really the key here. So what we started Portal to do is to develop the infrastructure for Internet businesses, because the network alone was not enough. The applications were going to be the interesting thing, but there was this gaping hole in the middle. So we were the doorway or the gateway to connect all these applications to the network -- that's where the name "Portal" came from back in 1985.

So we started writing the software, and about a year and a half later, we said, "The total available market for this software is precisely zero. We have a problem."

But what we found was that there were people, who even though they didn't want to buy the software, wanted to use it. They wanted to take their first tentative steps onto the network and roll out their first network online services. It was the same software underneath, we just kept refining it through version after version, generation after generation. This was really our beta program -- we had a lot of long-term paying beta customers telling us what they liked and what they didn't. We knew we never wanted to be a network company. Why compete for capital with an AT&T or an MCI?

That took us up to about five and a half years ago, and at that point the term "Internet" had been coined; the World Wide Web had been invented; Mosaic was available for free, and the number of users on the network was probably in the tens of millions. So we said, "OK, it's time to be a software company -- time to return to our roots."

We started to develop this product that we named "Infranet" -- infrastructure for the network -- and started selling it, and it's been explosive growth ever since. But really it's gone full-circle from wanting to do this infrastructure software, through the ISP/ASP/time-sharing service bureau years, back to being a 100 percent software company.

It's interesting -- I don't know if they've finally got rid of the few remaining lines of code that I wrote yet or not. But the trade-secret notice at the beginning of each line of code is exactly the same one we used 14 years ago. It's morphed a lot, and we know a lot more about how to build the software, but some of the initial design decisions are still there: It has to be massively scalable; it has to be real-time; there have to be lots of different ways to make money, lots of different business models; and it has to support this enormous explosion of Internet applications and businesses. We've seen nothing yet -- we're just like 5 percent along the way towards what it's going to be like.

CW: How is today's Internet different from what you had envisioned back in those early days?

JL: We didn't see the World Wide Web the way it's done now. We saw lots of information online, lots of ways to bounce back and forth -- the hypertext concept is more than 30 years old.

I'm not sure it's necessarily different from what we saw, but it was very fuzzy back then. It wasn't obvious what to do, because there was no market. And as the markets developed and people have started to build businesses and invest money and acquire customers, it's become more and more obvious what we need to do. I could show you the business plan that we used for the venture capitalists five years ago, and you could not tell the difference between that plan and what we're doing today -- the markets we're serving, the capabilities we provide, the general dynamics are all exactly the same. We thought about it for nine years, and we got something right.

CW: When you did your IPO in May, Cisco Systems bought 3 million shares. I understand that as part of that deal, Cisco is integrating Infranet into its Cisco Service Management technology. What's the significance of that in your view?

JL: It's a lot broader than just a particular technology integration. This was a decision that was made at the highest levels of Cisco -- (CEO John) Chambers signed the agreement; the board signed off on it; it's the largest cash investment that Cisco had ever made.

What Cisco realized, and if you find them in an off time they'll actually admit this, is that they are fundamentally a box company. They have great boxes -- they have routers and hubs and switches and all these boxes, and boxes are a very technology-oriented way of looking at the world. And their customers are starting to demand solutions.

Cisco started looking around and realized everywhere they were, Portal was.

For a small company in Silicon Valley, we had an amazing footprint, and we even had many accounts that they didn't have. Deutsche Telecom, for example, is primarily not Cisco.

So if it was just to do a specific product, there's no way we would let them buy any stock. They've made an unbelievable amount of money off the stock -- they bought 3 million shares at 13, and the stock's quite a bit higher than that now.

CW: Speaking of your stock price, do you happen to know what the ratio is right now between your market cap and your revenue?

JL: It's so ridiculously high that I don't even really track it. The stock market is a pretty interesting thing, because investors see the Internet is going to be an enormous phenomenon, completely pervasive, and people say they've got to be a part of this. But where do they put their money? That's an interesting one.

There's a lot of money chasing a lot of deals; everyone knows that most of those deals, most of those companies, aren't going to make it. There's a lot of shakeout. You look at all the previous industries -- the initial companies very rarely were the survivors. PCs: who survived the initial PC phase? Only Apple. Every other one of those early companies disappeared without a trace.

So people putting a lot of money out there want to know who are the long-term shakers. And the infrastructure space is a pretty interesting one. If you look at the Gold Rush of 1849 in California, it's the infrastructure companies -- Levi Strauss, Wells Fargo, Hearst newspapers -- that really did well. It wasn't the poor guys in the field with the picks and shovels and pans.

The infrastructure companies are very horizontal. If you look at Cisco, they can sell to a lot of different companies. And we're exactly the same. The fundamental things we do are: We manage the customers, who they are; we manage all the services that are provided to them; and we collect the money. So it's very horizontal; it works for Internet telephony, dial-up access, leased line, DSL, cable, satellite, wireless, online games, financial information, wireless PDAs -- an enormous range of businesses. We have customers today doing all of those things. So if one sector grows faster than another, we don't care, because we're underneath it all.

So, that's one reason that these multiples [of market capitalization over revenue] have gotten to levels that are just ridiculous by historical standards.

CW: Are they sustainable?

JL: Well, it depends. Two things have to change: either the multiples change or we grow into the multiple. And the last three quarters we've grown, quarter-to-quarter, 37, 37 and 35 percent. If that growth sustains . . . that stock is going to look cheap. If we fall off the wagon, then it's going to be very, very expensive. It puts a lot of pressure on us to really go out and achieve, but it's an enormous opportunity.

CW: Back to your IPO. In addition to Cisco, you also let Andersen Consulting buy stock -- 400,000 shares. What was the reasoning there?

JL: A very similar reason. Cisco said, "We think you're a great company. It doesn't make sense to buy you, but we really want to cement a partnership, and don't underestimate the signal that this investment will send, not just to the public, but to Cisco employees."

Andersen was a very similar type of thing. There's been some management turnover at Andersen -- (in September, former Managing Partner and CEO) George Shaheen went off to Webvan (a California startup offering groceries and other consumer products over the Internet), and Joe Forehand was promoted (on November 1 to managing partner and CEO from his previous position as managing partner for Andersen's global Communications & High Tech market unit).

Joe was the individual at Andersen who pushed this Portal relationship. We're, in his words, a "poster child" for how Andersen needs to work with these next-generation companies.

It was also to send a signal into Andersen that Portal is not like everybody else -- this is a very unique relationship. It was a lot smaller, stock-wise (than Cisco's purchase) -- it was only 400,000 shares. But it sent an unbelievable signal to the partners at Andersen that they were not doing business in the same old way, that this was the way of the future. And Joe's getting promoted, which happened after that, was unbelievably good for us.

Anderson has made a big commitment to teach us how to deal with the largest communications companies. Of course, we've been quite successful -- Deutsche Telecom was a deal we got last quarter. They're the second-largest ISP worldwide; first-largest in Europe. It's a deal we're working on quite closely with Andersen.

CW: To date, Portal has not made any money, correct?

JL: Right. We've been pretty much cash break-even for the entire history of the company, meaning the money coming in and the money going out are about the same. At the same time, we've been investing very heavily, especially in the past few years, in building out the internal infrastructure of the company.

If you look at the margins, the business is pretty sound. The software license business, we're 97 to 98 percent gross margin. You can't really get much higher than that. I think 97 is the same as Microsoft; we occasionally see 98. But we are investing very heavily in the future.

If you look at the cash balance, every time money comes into the company, it kind of stays there -- it gives us a cushion. It's not one of these situations where we're just grinding through lots of cash and at the last minute we get another round of investment and grind through that.

CW: Still, it's accurate to say Portal has never achieved profitability. How long can that continue to be the case before you start scaring off investors?

JL: The investors are giving us a lot of margin at this point -- we spend a lot of time on the road talking to them. But we have to keep achieving interesting strategic results. For example, I said that in the third quarter we're going to do three things: We're going to ship our 6.0 release on time; we're going to expand our international operations; and we're going to penetrate the high-speed access business. Three simple objectives. At the end of the quarter, we had shipped 6.0 on time; we increased international revenue to 38 percent of the total, which is very high for a company at our stage -- we announced deals with Deutsche Telecom and Asia Online and Shanghai Online and so forth; and we didn't just penetrate the high-speed access business. In the U.S. there are three big providers of DSL services -- all three are Portal customers now. And we got a bunch of tier-two players as well.

The market likes that. They like the fact that we set objectives and we meet them. As long as we continue that ability to hit those important marks and execute well, they're going to trust that we're doing the right thing. But you make a good point -- it won't go on forever. All of these companies have to go out and actually make some money. But that's how you judge a successful business in the long term. We're investing heavily for a 4X growth rate, which is about what we've had so far this year -- 400 percent annual growth is unbelievable when you sit back and look at it.

CW: Are you willing to go out on a limb and project when you're going to start to make money?

JL: With the laws on forward-looking statements, I can't actually say.

Expressing my own personal view would open the company up to all kinds of securities litigation. My personal opinion on that is that's actually pretty silly, because I do have a point of view, and we do have some idea of how we're managing the company. But if the slightest thing goes wrong, we get toasted.

And they don't just come after the company, they come after me personally, which is even worse.

CW: Your SEC filing stated that two customers accounted for a combined 28 percent of total revenue during the three months ended July 31, 1999. Is Deutsche Telecom one of those?

JL: Yeah, Deutsche Telecom was about 15 percent of total revenue. I don't think we announced the other one. We're in an interesting situation where we've got an enormous number of customers who say, "We don't want people to know we're running your software -- we think this is a strategic advantage for us." We might have four or five companies who are direct competitors [with each other] running our software, and none of them wants anyone else to know about it -- they all think it's a "secret weapon." But they're all running our software.

I should point out that for the three quarters before that, there was no one customer over 10 percent, so it's a pretty balanced business.

CW: What really cool things are you going to be doing in 12 months that you're not doing now?

JL: The cool things are not stuff that we do, it's stuff that our customers do. We're an invisible infrastructure -- if we do our jobs right, nobody ever sees us except for the people who buy our software.

The two interesting trends that are happening are Internet music -- the MP3 phenomenon -- and wireless Internet. The three objectives for the quarter we're in now are to continue our international expansion; deepen our penetration into the high-speed business; and start to penetrate the wireless Internet market.

Asia is a pretty interesting area for wireless Internet, as is Scandinavia.

The U.S. is a bit behind. But you can imagine, when the cell phone standards are enhanced to support wireless data, and the communications device itself is reduced to a single chip that costs a couple bucks, there's an enormous potential. It will be the applications that drive wireless, not the underlying technology. That's going to be an explosive market over the next five to 10 years. It's all going to be driven by business -- by making money -- and we want to be the infrastructure for all of that.

CW: So is it fair to say the R&D you're doing now focuses on wireless and music?

JL: The interesting thing about it is, the fundamental product is very broad -- for all these new applications, it actually works pretty well. For example, when we got into the high-speed access business, the DSL business, the in-stock, off-the-shelf software product worked for all those businesses, so we didn't actually have to do anything different.

So, we see the same thing with wireless. We'll have to do some technology innovations on that, but the fundamental product works very well -- it's not like we need a second product for the wireless business.

CW: How about Internet taxation? Are you doing any R&D there?

JL: We already have pretty good taxation capabilities because the communications companies have had to do that already. So we can support sales tax and value-added tax and a wide range of geographies. We've got an interface which plugs into best-of-breed tax software, so a company can pick whatever tax package it wants, and our platform already computes that.

The way we've approached a lot of these vertical specialties in general, like taxation or general ledger or reporting, is that we build an interface -- we build integration modules -- but we use other people's products to do those.

We want to stay horizontal and just plug into all of these different systems.

CW: What do you expect to happen with Internet taxation?

JL: Governments always like to tax stuff -- there's this insatiable appetite; it's kind of an addiction. If there's money there, they want some of it.

I'm actually fairly optimistic for the coming few years that taxes will be minimal in the whole Internet space, because governments do see an enormous potential here and they really want all these Internet businesses to grow. If they tax it too much in one geography, it will just move to another. That's the unique thing about the Internet, that you're not stuck in one location. If you tax it in New York state, it's all going to move to Connecticut -- it's a no-brainer to set a server up.

It's going to be taxed at some point. But I think there will be a reasonable set of taxation policies for these Internet businesses just because the governments can't get away with sticking it to people.

CW: Also with respect to R&D, Portal Software is an anchor tenant in Hong Kong's Cyberport. What was the attraction there?

JL: The infrastructure they're putting in place out there could be a really good thing for us. Being in a very stimulating environment with lots of other startup, interesting businesses is nice; the fundamental infrastructure -- the high-speed networking, the telecommunications.

We're moving into a new office building in California, and one of the big difficulties is that Pac Bell has had a great deal of trouble meeting their delivery schedules because of how much networking we need. We cleaned out two warehouses of cable; we put something like 800 miles of network cable in this building -- we're a very network-intensive company. If we can move into a building that's designed for that, that's great. It saves me money, but more importantly, it saves me time. Instead of taking six months to build out, maybe it's two weeks.

CW: When you go around the world and talk to your customers, what's the one thing they appear to be most pleased with? What do they say you're really doing right?

JL: An interesting thing that has developed in about the last six months is the maturation of business plans in the Internet business, whereas a year or two ago it was kind of, "If you build it they'll come -- we just have some idea and we're going to do it, then we'll figure out how we're going to make money."

In the last six months I've seen companies really starting to do business plans, and they're good, solid business plans. The thing that's the most pleasing is the fact that when they run their business plans through and they look at our software, we're not holding them up. They're doing all these new types of applications that we don't even know about. When the customer is delighted by some unexpected capability, then I figure we've done a great job.

CW: What's the one thing they appear to be least pleased with? What do they say you need to fix?

JL: I don't get a consistent set of things. I mean, people always like the software to be easier to use; everybody wants zero bugs, but they're pretty happy with the quality of the last few releases and how fast things are fixed -- we've put a lot of infrastructure into that.

I think the thing that they want, which is not something they can identify but is an interesting challenge for us, is they want us to stay ahead of their requirements. They're happy with where we are now relative to what they need, but they want to make sure we stay there. Deutsche Telecom has 3.6 million subscribers; we've demonstrated scalability far beyond that. But they just want to know that when they get to 5 million or 10 million or 20 million, the software will keep running. So, it's "how many steps down the road do we run ahead of them so we're in good shape?"

CW: In hindsight, is there anything you would have done differently?

JL: I look at the business, and (it's as if) after 15 years we're an overnight success -- I could have taken a nine-year vacation and started Portal. But I would say that thinking about this problem for such a long period of time has really given the company a very crisp insight into the Internet, where the market's going, how we need to participate. It's been a great 15-year ride, so I don't think I'd do anything differently.

Join the newsletter!


Sign up to gain exclusive access to email subscriptions, event invitations, competitions, giveaways, and much more.

Membership is free, and your security and privacy remain protected. View our privacy policy before signing up.

Error: Please check your email address.

More about AndersenAndersenAndersen ConsultingAsia OnlineAT&TBell LabsCiscoDeutsche TelecomGatewayGeneral DynamicsLevi StraussMCIMicrosoftPortal CommunicationsSECWebvanWells Fargo

Show Comments