Eric Schmidt became chief executive at Novell in April last year, at a time when the company was seeing $14.6 million haemhorraging from its profit sheet. He immediately slashed headcount and channel inventory, dropped more than 100 spurious projects, and is focused on Java networking applications. Schmidt is also leading the launch of the next major release of Netware. He spoke to Linda Leung at Brainshare Europe 98.
Linda Leung: What was your most daunting task when you joined Novell?
Eric Schmidt: Well I?d never been a CEO before and I also didn?t have a lot of financial restructuring experience. I had been told the channel was fine and there was no inventory problem and I was misled.
LL: By whom? ES: They?re no longer with the company. It was a mistake not to tell me the truth. If you had read the press you would have been aware of the situation internally. What was said externally was different from the truth. Basically I had to learn fairly quickly how to behave in such a situation. In retrospect it wasn?t so difficult. It was figuring out what to do and then doing it and doing it quickly.
The problem was the revenues numbers were not correct. There was too much inventory in the channel, so we had to take back inventory and we had to cut headcount by 1,000 people which was very painful. The best advice was to do it all at once which we did and we got it all cleaned out in the July  quarter - we?re in the July  quarter now so the quarter by quarter comparision will be quite outstanding.
LL: Where were the job cuts made?
ES: All over, but general administration was hid the hardest. The company had administrative staff suited for a $2 billion company, we?re only $1 billion in revenues. I said 'why don?t you get rid of all that?' and they said 'we were expecting revenue to come back', which was stupid. It?s not good thinking.
LL: So a year later, what is now your primary goal?
ES: The important thing is to pause and say viability is nolonger an issue. A year ago people were saying the company is going to be for sale, it?s going out of business. You don?t get that any more.
LL: You do occasionally.
ES: Well, they?re wrong. Nobody who is in the know says any of these things.
LL: But if the right buyer came along...
ES: It?s not for sale. The answer is no.
ES: Because I said no. It doesn?t need to be for sale. There are companies that are smaller than Novell, buy those first. Novell is a huge company it doesn?t need to be for sale. We?re not for sale, we?re not being bought and we?re not planning on buying anybody - we?re big.
LL: Are there are gaps in your product portfolio you?d like to fill with an acquisition?
ES: There?s a lot of stuff I?d like to do but prices are too high and I?m cheap - I?ll wait until their prices come down and our value goes up and we?ll make an even better transaction.
LL: Which kind of technology would you be interested in?
ES: I?m not focused on it. I?m trying to get our products out the door.
LL: How do you rate the importance of Netware 5 to the company?
ES: Netware 5 is important to our company because it establishes a baseline platform for TCP/IP leadership. Up until now we have been an IPX company with an IP gateway, which is not the same thing as being native or pure IP. The Netware 5 kernel has virtual memory, it supports symmetric multiprocessing, it has fastest IP engine on the planet. There are a lot of reasons why Netware 5 is incredibly important.
LL: When Netware 4 was launched many people stayed with 3.x. How can you be sure Netware 5 will not experience the same fate?
ES: [The number of customers upgrading to Netware 4] was slow because it wasn?t compatible with 3. The first rule of installed bases is don?t break compatibility. So Netware 5 runs all Netware 4 applications so it should be a straighforwad upgrade. It?s a little bit easier to go from 3.x to 5.0 than 3.x to 4.0, but not much. So basically if you?re considering moving, you probably want to just go to 5.0. But frankly a lot of people are happy with 3.x and they?re not going. We have brought a version of 3.x for those customers and they should be happy for many years.
LL: Is it wise to launch a big upgrade when customers a have other worries: Year 2000 and Emu?
ES: Our customers have been waiting for Netware 5 for three years - they?re dying for this product - we?re late. I?m not worried about adoption, I?m worried about making the product work.
LL: When will it ship?
ES: Sometime in the summer - its current schedule date is July. It won?t be sooner than that.
LL: How are you addressing Novell?s market image, compared to Microsoft?
ES: Microsoft is a not a comparable example. Microsoft is being sued today, is that a good market image? A comparison to Microsoft is not fair - Microsoft is a monopoly, just as IBM was 25 years ago. Microsoft controls 90 per cent of all the clients, more than 90 per cent of the office, and a fair amount of the servers. We?re not Microsoft and nor are we going to be Microsoft. We fix the problems that Microsoft creates. If you?re an NT user and use NDS you get security and a directory you can?t get from Microsoft.
We?re a software company that fixes some of their problems. Also, Netware doesn?t compete directly with NT. Clearly executives at Novell spent a long time saying it did - it doesn?t. There?s a whole bunch of things that Netware is not good at, which NT is good at - they do two different things. You use Unix if you want large complex databases to work; they don?t work on NT. The directory stuff doesn?t work in NT, if you want a directory you have to use Netware. We have much lower total cost of ownership for Netware and NT together than NT alone. So the message is mix NT and Unix; you can?t achieve business objectives if you use just one.
LL: You said you are a software company - what happened to being a networking company?
ES: We are a networking company. I use the terms software and networking interchangeably. Novell is the primary largest marketshare networking software company in the world and we?re gonna stay that way.
LL: Do you still want to be a Java server company?
ES: Sure, but those applications will be network centric. Java is a language. The term 'Java app' means nothing; you have to say 'It?s a Java app to do networking', and so on. Java itself is like C, so when we say we are a Java applications company, the correct technical term is to say 'We?re a Java server networking applications company'.
LL: You?re confusing the market, because the industry thinks you mean general applications rather than networking software, which during this conference Novell has termed ?specialist? software.
ES: I appreciate that - that?s a problem we have. You would use us as part of a Java solution on the server if you want scalability and performance, and because of 'write once run anywhere', those Java applications will also run on Unix and NT, but they?ll run faster and more cheaply on Netware
This is the distinction - do you believe Oracle is a Java app or a general purpose app? It sits on the network, it comes up with SQL, ship it down the network, answer comes back through SQL. The conventional wisdom is that?s a general purpose app. It isn?t, it?s a specialised network service which just happens to be in SQL.
Part of the confusion is there?s a new model here. Microsoft?s model is a thick heavy client, thick heavy server, and confusion as to where the role is. The correct model, from a network science perspective, is to have a relatively thin client speak a protocol, of which SQL is an example, to a server which is managed by adults, or at least engineers.
Microsoft?s whole model depends on you as a personal administrator to your desktop which increases the cost of ownership and the security risk within your network. We?re pushing a different model, called an executable content model, where there?s a set of servers which has protocols, and the desktops talk to the servers in a scalable way.
LL: When do you think Java will be in business critical applications?
ES: It?s on schedule, it?ll take another year. It took 15 years for C, and 10 for C++ to get this far, so we?re two years away for Java. There?s quite a few interesting applications from start-ups. The major software developers are not going to use Java except on the peripherals because they already have code. It?s the start-ups that will be innovators in applications.
I?m an investor in something called the Java fund, as is Novell, and we have some enormously interesting companies that are doing software metering, Web site management, electronic commerce, all using server-side Java.
LL: What?s the reasoning behind Novell giving away as many sourcecodes and interfaces as possible?
ES: It?s pragmatic. In our world we have to interoperate with everybody, including Microsoft, whatever it takes to make sure we interoperate. Our customers insist we work well with Microsoft. It?s not an option not to. So if they ask for Active Directory [Microsoft?s network directory server in NT 5], we?ll support Active Directory. We already support LDAP.
LL: If Novell didn?t have such a sizeable installed base would thing have been more difficult?
ES: Oh, very much so. I can tell you as a very recent expert on turnarounds, that turnarounds require bigness. You need a lot of cash, you need a big installed base, a lot of employees, and a lot of customers.
Because you have to throw the switch really big, you can?t go half way. Our cash position went down by $100 million which for any reasonable sized software company is an immense amount of money. It came back a quarter later, more than $100 million, so we said 'Oh, OK, $100 million in, $100 million out'. We?re getting a tax refund this week of $69 million from the US government. So having bigness helps in turnarounds.
LL: But bigness also means a bigger fall if things go wrong.
ES: Of course, but they didn?t. What?s the alternative? Do you want everything to be Microsoft - you don?t want any choices at all. I think it?s important that Novell, Netscape, Apple, IBM, Computer Associates, Oracle, Sybase, all of those companies need to prosper. We need competition.
You don?t think that Internet Explorer 4 will be any good unless Netscape is such a good competitor. You don?t think Active Directory will be any good except it is trying to copy what NDS does. Competition drives markets - you need a competitor. In Sun?s case its primary focus seems to be to compete with Microsoft these days. You don?t think that has an effect on the quality of NT 5? I?m sure it does. NT 5 is going to be a better product because of Sun?s competition. It?s also a later product.
LL: Did you receive any advice from Scott McNealy [Sun?s chief executive and Schmidt?s former boss] when you joined Novell?
ES: Sure, he said don?t go. But he also said don?t blow it. Scott is a talented CEO and during the 14 years I?ve worked for Scott I learnt a lot and it came it very handy so far. He?s a very smart, and very shrewd man. He is very focused on executions and is financially conservative. Guess what, I learnt all that from Scott. I think he?s pleased with my progress. He?s impressed.
LL: What would be your three pieces of advice for businesses in the next century?
ES: Well, first they need to get to the next century. The world is going to get networked to a degree that is unfathomable today. The telecommunications networks will be the primary network in how you reach customers, do business, and communicate with your employees. It?s reasonable to say the successful or failure of a business is going to be how network centric the business is run.
Statement two is that unfortunately there is no single vendor, or standard, or solution that will deliver all those for you. You will have to purchase best of breed. If you choose to standardise on a single vendor you will miss out on what another vendor is doing and that?s called competition.
The third issue is for a European audience. Europe is behind with respect to online adoption. The percentage of Europeans, with the exception of Scandinavia is at least half that in the US. This is due to out of date regulations. This is a very serious issue long term. Any gap that exists in online usage between Europe and the US is very bad for Europe. What needs to happen is regulation, which is very painful. This gap could be filled very quickly if the leaders had the will to do it.
Some parts of Atacama have not received rainfall for 500 years - but a sudden deluge of water upset the Desert's delicate biological balance
Spitzer Space Telescope could not spot Oumuamua, suggesting that it is actually pretty small
Greenland crater one of the 25 largest impact craters on Earth
This long-sought progenitor star was identified in an image captured by Hubble in 2007