What does the 600 million dollar man look like? The door opens to Jeet Singh's hotel suite. He is wearing a slightly crumpled green shirt, a pony tail and the weary expression of a man who spends too much time on a private jet.
Singh, CEO of e-commerce software company Art Technology Group (ATG), was named by Forbes magazine as the 23rd richest man in the US under 40 years old. He is Indian by birth, was "educated in the Himalayas" and got a politics degree from Massachusetts Institute of Technology (MIT).
After a high-level career in IT marketing, he teamed up with fellow rock group member Jo Chung in 1991 to form Art Technology Group.
Their e-commerce platform Dynamo is now head-to-head with big names like Broadvision in the e-personalisation market. ATG went into profit two months earlier than expected and is hiring "100 programmers a month" to meet growing demand.
Personalising the Web is the key to Singh's e-vision. But is Web personalisation just a fad driven by the need to add bells and whistles in an overcrowded B2B exchange market?
"I would describe the marketplaces industry today as looking very much like e-commerce looked two or three years ago. Then, everyone was concentrating on transaction. The value proposition of a market exchange is very thin - because it's not hard to do.
"When storefronts were built three years ago, there was no connection to the marketing site or customer service. The customer relationship management (CRM) market is interesting because now these parts of the business are finally being connected.
"B2B marketplaces are similarly isolated today. The people who put up the exchanges have no connection with others in the same company who are building customer management.
"The most sophisticated users are finally starting to realise that they're managing a customer segment. There's a moment when they are working in an exchange and there's other times when they are not. The front-end of customer management needs to be connected at various points to an exchange or a procurement system, even other knowledge management systems."
Singh also insists that, at present, B2B exchanges are lowering costs only in the low-value segments of vertical markets. "You may see Boeing and Airbus lowering costs through an exchange, but they are just reducing the cost of nuts and bolts.
"They won't share an exchange for the top 20% of their cost basis as that is their competitive advantage. They're mostly operating like a cartel now. When you see the big exchanges being put together, they're just trying to cram down the prices of their cheapest suppliers."
Beyond cost reduction and supply aggregation, how is B2B e-commerce changing the internal business model of user firms?
"The idea of creating products becomes so much broader," says Singh. "The products you create do not exclusively come from within your four walls and then suddenly get taken out to the public. Today in the car buying or technology space, you source things from other vendors. Then you package them and sell them, right? I think you will see more of that in a lot of other industries. So the idea of packaging services, ideas, concepts, knowledge, and redistributing becomes a core part of business.
"There is a more interesting, bigger trend: customer segment return on investment (ROI) management. Right now, profitability is measured by department in large organisations. It's a very bad measure. Marketing is measured by how much is spent and how many leads are generated.
"With IT infrastructure, you don't even manage ROI because, say, it's an SAP project for five years costing approximately $100m and you never knew what came out the other end. You're just hoping for the best.
"On the other hand, customer support is very easy to measure - $20 per telephone call versus 20 cents on the website. But each of these areas cut across a customer segment. College students, retirees, life assurance buyers in financial services, for instance.
"The most sophisticated firms are trying to understand profitability by customer segment. You don't necessarily aggregate all your costs. You take 20,000 calls in your call centre a day? Well 10,000 of those calls were from people who buy your $5 product. It's not worth taking those calls. You put it on your Web site. The other 10,000 are people who are valuable. You should be calling them first. Spend more money and take care of them.
"If you think of customer segment value, it changes the way you create new products. In the past, people have created products by adjacency. You build ovens. Your factory can now build toasters. Now why don't they sell mutual funds or insurance? What we own is the customer, not the product. Instead of a car company buying up a motorcycle one, you might see a car firm acquiring a financial services company."
Segmenting customers is all very well - but won't it actually increase the digital divide? Won't the bottom 50% get left on the digital scrapheap?
"I think you are right. I remember the early days of the internet. Everyone said if it has great communication capabilities, people would be much more open to other cultures. Things like racism and prejudice will die down.
"I was putting the counter-argument. It's easier for all the neo-Nazis to get together anywhere in the world.
"The internet is a neutral medium. I think you seek out enhancement on both sides. I see the effect in my country, India. The internet allows people who normally would never have jobs to work at pay scales that are starting to come close to the those in the US. It's amazing. How could that possibly happen?
"But if you look at India as a whole, what is being described is absolutely true. There are people who are amazingly rich next to people who are amazingly poor.
"In the US, it is the same. There are people who have lower skills and their pay is dropping. The people at the top are getting richer and richer.
"I think the policy of inclusion is a good one. But I don't think you can create social equality that way."
As a growing software company, ATG obviously has its own skills problems. How is Singh overcoming this?
"We are surprised we can actually recruit at the rate we are. But we have a lot of things going for us. Our reputation is high. The industry is hot and there are people who want to work for us.
"The kinds of incentives people are giving for hiring are really ridiculous. They give away Porsches. Obviously salaries and stock options are high. Now it's benefits like the work environment and massages. It's nutty, like a price war.
"You can't give away stock forever. Stocks don't always go up. Price wars operate against you. At the end of the day, how do you differentiate yourself?"
ATG's software is built on, and runs through, Java. It is the biggest skill in demand. Isn't that a problem? Singh thinks not.
"It's also the only popular thing you could hire people to do. If you're not going to offer a job in Java, it's very difficult to hire someone. If you're a C++ or a Cobol programmer who just came off Y2K, the last thing you want to do is get another Cobol project.
"They want to get new skills because they know that the salaries are going to be better. Their demand is going to be higher in Java.
"You can't recruit a person out of university and offer him or her anything but a Java job. This is why Java has taken such precedence now. There was a question mark a year ago, now it's gone."
Many employers in the UK are tearing their hair out over C# as they can't plan their skills strategies. As a Java champion, Singh is predictably scathing about Microsoft's Java-busting new language.
"I think if Microsoft really proceeds down that course, it would be a huge mistake. I find it almost unbelievable. I don't think it will be that close a battle. I still don't know what's going to come of it. I know Microsoft have been planning a Java release at the same time. We know people who are at Microsoft building it, so it's interesting. But I would say even Microsoft shouldn't be fighting at this point."
What will the internet mark II look like?
"I think it means heterogeneities. It's going to be bandwidth-heterogeneous and device-heterogeneous. That's the only thing that I can possibly predict about it.
"That has huge implications for the issues vendors and users will have to confront. Developers have to confront building applications in terms of usability. That isn't so easy. It isn't just a browser to run a PC at a bandwidth X that happens to run be a little faster if you're on a T1 connection.
"Everyone talks about broadband like it's going to be very different. What people mean by broadband is it's going to work like in your office. T1s are not exciting. But heterogeneity in applications is very complicated.
"You've seen what interfaces look like on wireless application protocol (Wap) phones or personal digital assistants (PDAs). They can be tough, clunky and very problematic."
So what does the new environment mean for personalisation? "When we do a wireless device or a website, we think 'what device is it?' We send different data. As a technology problem, it's not complicated.
"Personalisation is a fundamental solution to multiple devices. You have an interactive television, you're on a PDA - this is what we send you. It's just like segmentation, but you are segmenting content. This is what we do anyway.
"Put it this way - it's not a technology problem. I think it's a usage and design problem. Not enough work is happening in usability these days. People gave up on usability and now they've said 'to hell with the Net, it's got to be fast and work'.
"I hope companies which have experienced usability designers can get back in the act because we need them as well as customers.
"People are just disposing of sheer utility, which I understand, but with no regard to the limitations and possibilities of the device."
Today's Web usability benchmarks are very much based on the screen. Does Singh think there is an emerging consensus about what makes good usability on the Net over many devices?
The short answer is no. Too many design companies are living for today's cash flow. The heavier programming companies have been swept away by the services boom, he says.
"On the other hand," he adds, "as new devices come in, new skills have entered the market. Now you have people who are Nokia phone or PDA phone designers. They say 'we have a new skill-set here and know something about how people operate. This is what the application should look like.'
"I don't think the killer app is going to be a brand new technology. The killer app is going to be when someone finally figures out how to make a heterogeneous internet operate in a way that makes sense.
"It hasn't happened yet. There are some things that are simple that people use a lot. Stock quotes on a PDA: big deal, that works - thousands of people use it. But it's not a revolution."
What about Microsoft's .net strategy? Isn't it a paradox that Microsoft, so vilified for the way it treats customers, is the one big vendor with an holistic strategy for the user experience on Internet mark II?
"I think the people who are really in the position to do it are telcos. It's the service provider who is in the best position to do the aggregation. Why? Because that's the role. They're the ones with access to potentially different services that could be aggregated for the user. At the same time, they could have that skill set."
Singh thinks that the contenders in the market who could create an holistic user experience are AT&T, Sony and AOL. "Microsoft is actually the least likely, I think, because it is the most product-oriented. It is the least user-oriented of the groups. Large media companies which also have a collection of assets in TV and are now getting into broadband are also in the running. That's why I said AT&T."
Who does Singh think has really understood the potential of the internet in today's marketplace?
"Financial services are the most sophisticated by far. They are high margin businesses that can afford to invest in technology. They can afford to spend money for a return five years up.
"A retailer can't. They have lower margins and can't invest in technology. They never have invested in it, so they're not even used to it in the first place. Even their accounting can't deal with it. Those are the extremes - retailers on one end and financial services on the other.
"Retailers are very sophisticated about customer segmentation and management. They don't have money to invest, but they have a lot of knowledge.
"Financial services are very savvy about segmentation. They understand the difference between a university student opening a bank account and a mutual fund owner that yields thousands of dollars a year.
"Then there are companies like telcos which gain high margin and investment, and have pretty good demographic understanding. They do segments demographically.
"I believe telcos are an oddity in that they have money to spend capital intensively, almost prior to even knowing what service they'll provide. This is interesting. It becomes even more apparent with the huge investments in 3G licences."
From Singh's vantage point, what should IT directors be worrying about? What should be top of their agenda in terms of the e-commerce road map?
"The number one thing they should be worried about is future proofing. They cannot afford to put up a three-year roadmap in a lot of ways because the chances of them being right are slim. That's a tough position to be in. It's harder to plan shorter term. So they have to be worried about flexibility. I think that's key because if you don't have it, you just can't correct your errors.
"In a chaotic market, one that hasn't really settled down and matured, the chances of being wrong are really high."
Singh believes the uncertainty is what contributes to the widespread support for Java.
"The level of support surprised us. We chose Java for technical reasons. We thought it was the best and cheapest way to go. But we didn't realise or know what an IT manager of a big company would look for six or seven years ahead. We weren't selling to those then.
"The IT manager now is looking at a budget of $10m over the next two years for e-commerce or $100m if you are GM. It's a big number. Of those dollars, half will be spent on services. So you'll spend 10% to 20% on applications. You're going to buy Sparc stations, Oracle databases . . . and then you're going to spend 50% of your money on Andersen Consulting.
"That custom code might eventually have to move. Maybe it's to an NT platform. So they want to buy-in Java because it could be ported if, at worst, their entire infrastructure changes. That is a future proofing decision."
The second piece of advice from Singh is to focus. "Most IT managers face a very big feature list from their CEO, so they are trying to prioritise. Instead of doing everything at once, I think it's extremely important for IT managers to make sure they do the one thing that most people care about first. And get it right. That usually means build a 'platform'. In other words, build something that will please many of your constituents quickly and then build up the rest.
"Now we're seeing budgets consolidating within large companies. CIOs are starting to reel them in and are saying: 'There are 40 projects going on and you've chosen 40 different software products. There's a 50% overlap. You're all building the same thing. Why?'"
t Next week we talk to Web usability guru Jakob Neilsen
JEET SINGH: CV
CEO, Art Technology Group
Jeet Singh co-founded Art Technology Group with fellow rock group member Joseph Chung in 1991. Prior to founding ATG, Singh was a marketing and business development executive at firms including Boston Technology, Team Technologies and Groupe Bull/Bull Corporation of America. He received a BSc in political science from MIT. He was listed by Forbes magazine as the 23rd richest American under 40.
Singh on the anti-cookie backlash
Aren't companies that base their Web strategies on personalisation taking a big gamble? Won't there be a consumer backlash against cookies and site visitor tracking technologies?
"Privacy is being abused in exactly the same way as in the past. In the US, I think there already is a backlash. But I think it does point to the fact that you are putting sophisticated tools in the hands of inexperienced users. We've invented guns, but we haven't invented the safety catches yet. There's definitely that happening in the market," says Singh.
"A lot of companies in our space are fairly sophisticated when using customer data. What they are trying to do with the Web is actually use it in the same way as in the past. In some ways, there's the 80% of the market using sophisticated tools to do something that was already sophisticated. Also there is 20% of the market that have never really done anything like that, who are going bananas with these tools. But they don't really know how to use them - as they've never done it before."
Singh on the IT skills crisis
A big issue in the US is H1 visas (importing skills to the US). "Can we import skills or not? A lot of people ask me: 'Isn't this fundamentally an education problem?' Yes, but there's a 10-year lag. That's school plus university. By then, not only will it be solved, it will be the other way round. Remember that we have had a glut of lawyers and doctors in the US. Every time there was a shortage, everyone went to school saying, 'I'm going to be a doctor'. Ten years later, they all come out and there's no jobs left. The system does normalise itself in the long-term."