Reinventing NeXT

Steve Jobs goes on the record about technology, marketing, and Ross Perot

By May 1992, NeXT had completed its transition to a marketing strategy based on object technology and custom development. Steve Jobs was by turns evangelical, combative, and reflective as he sat down with NeXTWORLD Editor in Chief Dan Ruby for a wide-ranging discussion.

NeXTWORLD: It's been six years since NeXT was founded. Are you where you thought you'd be?

Jobs: No, I would say we're behind where I thought we would be. Not so much because we're much different than a lot of companies, but I'm impatient. I think we're like a year behind where I'd like to be. We changed the whole strategy of our company last fall. We changed our marketing strategy. We changed our sales focus. We've been focused consistently on mission-critical custom apps for a little over six months and it's really working.

Why is custom apps such a potent idea?

Getting applications written is the number one problem in corporate American information technology. And even though we're downsizing to client-server computing and the apps are on the desktop, the bottleneck is still getting them written.

This is even more true when customers want to use computers for operational productivity as opposed to management productivity. You can't buy shrinkwrapped software to do stock trading or run your hospital or do order processing. You've got to write custom apps. Now, in the past, these operational applications were written in COBOL or some more modern language on a mainframe or minicomputer. Starting in the very late '80s, some companies started downsizing to client-server computing. They could buy a Sun and spend like two years writing a good app, or as good as you could write on a Sun. Now, we roll in and say, look, you can write that custom app five to ten times faster on a NeXT.

Who is writing all these apps? Is the market large enough to support a company like NeXT?

It's incredibly important in some markets. Take financial services. When they roll out a new product it's only three things: an idea, a sales force, and a custom app running on a desktop banging SQL databases on a server. Without the app there's no new product.

Take health care, which is starting to explode. People in that industry are starting to realize they have to automate the core operational activities of that business. Or legal, which we think is a very ripe environment for writing some custom tools.

These are vertical markets, but we view custom apps as a horizontal benefit in the same way that desktop publishing was in 1985. Eventually, everybody is going to realize they need to have mission-critical custom apps.

The custom-apps strategy tends to de-emphasize the need for third-party apps. No. I disagree. There was some concern at first on the part of our third-party developers. But we're finding that it's actually helping to sell more apps. When Apple focused on desktop publishing, Microsoft sold more spreadsheets because desktop publishing was the Trojan horse that got the Mac into corporate America. So mission-critical custom apps is the tip of our arrow, getting us into these major corporate accounts.

Some people have said that custom apps is a really smart strategy that gets NeXT's foot in the door, but that it needs to be followed, or supplemented, by some other approach.

Well, I would say two things. Number one, we didn't make up mission-critical custom apps. Our customers told us this. So this is a strategy that's being generated by the marketplace, not by some marketing person in some secret room at NeXT. We've hit bedrock. There's a fundamental need that's about to explode out there in the marketplace that requires object-oriented technology. And we've got a three- to five-year lead on any competitor.

Now, does that mean mission-critical custom apps aren't going to get supplemented with other things? Of course not. We have the best collaborative-computing environment in the world.

What you used to call interpersonal computing (IPC).

You could give it any name you want. You could call it groupware. You could call it interpersonal computing. You could call it collaborative computing.

What do you call it these days?

Well, that's not the biggest problem. The problem is that people are not running around thinking they have a problem. In other words, I don't run into customers who are pulling their hair out saying my collaborative computing environment isn't good enough, not the way people are saying they have to get their custom apps done faster.

So we have this wonderful thing, this collaborative environment. It's the PC that everyone hopes is the PC of the mid-1990s. We've got it today. And when people use it, they love it. Even our most custom-app oriented customers fall in love with the interpersonal computing environment on the NeXT.

But the issue is, how do we communicate how great it is to people before they use it?

We aren't alone in this. Lotus is also having a hard time articulating what Notes is, and the market is having a hard time understanding it. So this whole idea of collaborative work is a very new thing and the marketplace doesn't have a language to talk about it yet.

But in terms of future messages, the IPC story is still viable?

Oh, it's real. It's absolutely real. Different customers buy our computers somewhere along a spectrum. Some are weighted toward the custom app and they're interested in the collaborative computing. Others buy it for collaborative computing and they're interested in writing some custom apps someday.

Let's shift away from marketing and look at the business. In terms of numbers, sometimes people throw around the figure of a billion dollars as a milestone to be reached. What's your timetable?

I don't think in those kinds of terms. I think in terms of $200 million, $250 million, $300 million. You know, we did $128 million in '91 and I'd like us to produce some healthy growth this year, which I think we will.

There's been a lot of change in the management structure at NeXT. The most current one is you've just brought in a new COO.

Sure. There have been four big changes in the last six months. The formation of the Hardware Division with Rich Page running it; bringing in Marcel Gani as our CFO; promoting Mike Slade to vice president of marketing; and now we've brought in Peter van Cuy-lenburg as our chief operating officer and president.

What I want Peter to focus on this year is making us operationally more effective. We have a very large investment in our sales engine, distribution strategy, and manufacturing. Those are the kinds of things Peter's very skilled in. This is freeing me up to concentrate more on things like our product and marketing strategies. Okay, let's talk about products. Is NeXT a hardware or a software company?

I would say that the introduction of NeXTSTEP '486 is clarifying, both inside NeXT and for the outside world, the strategic model that we've always had. We're a software company that happens to make great hardware. We have never for a moment thought that our true competitive advantage ultimately wasn't software. It is. So NeXTSTEP '486 will come out, and who knows, maybe other versions of NeXTSTEP will come out. I think NeXTSTEP '486 is going to be very successful. We're getting help from a lot of companies from almost every facet of the PC world. They want to ensure that there is a choice and they see NeXTSTEP as a really good choice.

When should we expect to hear some word on who will be supporting it?

A lot of things are already completed but we don't feel a need to announce them right now, and we can get an even broader base of support in the meantime.

Will NeXTSTEP 3.0 be out by the time this interview is published?

Well, we were about two to three weeks late going into beta, which probably means shipping it may slip into early July.

With all announced features intact?

The government may not let us export encryption. But everything else is there.

Of all the new features in 3.0, which ones do you get most excited about?

Well, DBKit would have to rank number one for the customers. It's a breakthrough technology that's only possible on top of our object-oriented framework. Number two is probably the integration with Novell and AppleShare, which provides bridges to our customers' existing environments.

As for my own point of view, I believe that we've made incredible strides in color. Also, the object linking and distributed-object technology is profound. We're several years ahead of Apple, Micro-soft, and OMG (Object Management Group) in distributed objects and object linking. And while that isn't going to be apparent the first day we ship, because apps that take advantage of it are yet to be written, a year down the road that may well emerge as the most profound technology in 3.0.

The 3DKit is also going to be interesting for business presentations and some very exciting CAD stuff. It is also going to take a year to blossom in terms of the apps, but I expect a lot out of 3DKit.

When you introduced 3.0, you gave the impression that we really needed more powerful machines to take advantage of 3-D technology.

Well, that's not true for photo-realistic rendering of single frames. As for Interactive RenderMan, you're correct that to really challenge the Silicon Graphics (SGI) of the world, we'll need the next generation of hardware. But a lot of apps are going to get written on the current hardware, so we've gotten the development process started a little sooner.

Now that we are talking about future performance improvements, what can you say to users about what they might expect?

Over time we'll broaden our hardware line to include, in addition to the current products we ship, computers with even greater performance. That's no secret. These products will cost more, but I don't think anybody will have anything on us from a performance point of view.

You have somewhat limited pockets to develop or acquire lots of different kinds of new technologies. How do you stay competitive?

The most important advanced technology that we're working on is object-oriented software. It's far more important to us than most of the other sort of sizzle that people talk about. You have to choose carefully what you invest in and then stick with it. For us, object-oriented technology is at the top of the list.

What's happening is that companies are now selling objects for NeXT. We're focusing a lot of resources there because we believe very strongly that this is the future. NeXTSTEP is the only software platform that can really support the creation, sale, and use of other people's objects. That is very important in the competitive battle as well as in helping customers in creating a whole new subindustry.

Your competitors say that NeXT objects are not standard that they aren't portable to other object environments.

They are portable to NeXTSTEP '486 and any other version of NeXTSTEP. As you know, we've shipped around 50,000 object-oriented computers, and right now that's more than everybody else times five. So right now we're the standard, and OMG simply addresses some very low-level issues. Other object standards will appear as other people start shipping object-oriented computers, which we don't see happening any time soon.

It's interesting that you've identified this custom-apps market and suddenly we see other companies also talking about that.

Taligent has almost copied our marketing literature, which is fine.

When do you expect a Taligent operating system to be released?

In 1995. They might release some beta stuff in '94. And we suspect very strongly that the first release of something will probably not even be as good as NeXTSTEP is today and will have to be tuned. That's assuming they even ship a product.

You have some doubt that they will?

Well, the last thing [Taligent CEO Joseph] Guglielmi worked on was Office Vision, when he spent $300 million and never shipped anything. So who knows what will happen. I wish him the best.

Microsoft NT seems to be targeting custom apps as well.

No, NT is very different. NT is just Microsoft's second attempt at a UNIX wannabe. It's better plumbing for Windows, but you're still stuck with the Windows development environment, which, if you talk to developers, most say is the worst development environment ever created. So I don't think that you're going to see NT going after the custom-app business.

Remember, the actual operating system is only 10 percent of what NeXTSTEP is. Operating environment is a better term. In order to set object models and really be a player in it, you've got to do the other 90 percent. So NT gets you 10 percent of the way there. I don't see that changing very much over the next few years.

Even as they develop higher-end systems?

We're not seeing it happen. NT will offer better networking eventually, but the whole PC world is locked into Novell anyway.

Microsoft seems be developing a relationship with DEC that looks like it might be more formidable than the ACE coalition would have been.

If you believe that DEC will be more formidable in the future then you would say that might be true. It's hard to say what's going to happen with DEC.

People are very excited about their Alpha technology.

We're thoroughly familiar with their Alpha technology, and I see it in the same ballpark as HP's technology. If you look at the real performance of Alpha, it's good but it's no better than what HP and others, possibly including NeXT, have in store for 1993.

Okay. The mother of all competitors is Sun Microsystems. Let's discuss this war of words or whatever it is that you've got going. Some people think that you picked the fight.

Oh, absolutely not. This is far more than a war of words. Scott McNealy is keeping the airlines in business. Every time we're about to close a deal he flies out and tries to talk the customer out of it. He did it with the L.A. County Sheriff. He did it at Phibro. He did it recently at Citicorp, where we just won a big order against Sun. Sun is trying to stop us with everything they've got. Fortunately, they're not succeeding because they don't have a competitive product.

Has it reached the point of personal animosity?

The slander sheet they put out was a personal attack on me. We don't do business that way. But I have offered to debate Scott McNealy any time, any place, if he wants to talk about the issues.

In the meantime, NeXT has been pretty aggressive itself.

We are absolutely going on the offensive as a result of their actions. We are not standing still over this. We watched Sun put Apollo out of business and Apollo had better products than Sun, so we know that we have to act and not dismiss this as unimportant.

How about Apple? And not as a competitor, but from your unique perspective. How are they doing?

Well, I think that anybody who's managing to survive and grow a little bit in this environment is doing well, so I would say Apple's doing okay. I think that Apple wants to be the consumer-products company for computers. As an example, PowerBooks are very good Macintoshes but the Quadras are terrible. So it's obvious that Apple is putting its best people on the low end of their line, and their high end is withering. Look, they basically sold Taligent. They took Pink, which was an internal Apple technology, and basically threw it out.

Well, they spun it off into a new company.

Right, which is not what you do with your company jewels. So from my point of view, Apple was at a fork in the road for a long time, and they finally chose a path of becoming the Sony of computers. That's probably an excellent choice for Apple, but it means that there are going to be a certain number of people at the high end of Apple's market that are over time going to feel disenfranchised.

One of the interesting things as we look at the development of future operating environments and advanced systems is that there are a lot of coalitions that form and then sometimes break up as well.

Right. ACE has fallen apart. The only people seriously using MIPS processors is SGI. So SGI just bought itself a processor division.

In general, what's your feeling about alliances?

Well, name one really successful product that came from an alliance.

How about MS-DOS?

No. That wasn't an alliance. MS-DOS came from Microsoft. I can't think of one successful product in the history of our industry that ever came from an alliance. Or any other industry, either. Toyota and GM no. I'm sure there are some examples but they're not standouts. So that's my answer. That's also why Taligent may never ship a product.

We work together with tons of people. We work together with customers, we work together with developers, we even work together with competitors sometimes. But the statistics are not in favor of these industry alliances.

Let's expand this discussion beyond NeXT and the computer industry. A lot of people are interested in your thoughts about Ross Perot and his presidential candidacy.

I think he's got a real chance and I'm helping him every way I can.

Such as?

I'd rather not go into details. But I think Ross would be an excellent president.


Leadership. He chooses very good people around him, and he is able to distill out of his troops and himself a clear vision. Then he's able to articulate that vision well to a large number of people.

What's wrong with our country is that we haven't put the energy or resources into developing our collective values. We have to take some time and ask, "What are the ten most important things that we want to do as owners of this country?" and then go figure out how to do them.

For as long as I can remember, the only real leader that was able to communicate a vision was John F. Kennedy. I remember watching him on TV when I was six, seven, eight years old. When he said let's put a man on the moon and bring him back in this decade, that was a vision that a scientist, an historian, and a gas station attendant could understand, each in his own way.

Let's finish with one personal question. The big change in your life in the last year, aside from custom apps, is that you got married and had a son.

We had our own little custom app.

Has that changed your view of life or work?

Having Reed has been a different experience for me that I won't even attempt to put into words. The only thing I'll say the only thing I really think I can say is it's as if I never saw the color green before and all of a sudden when Reed was born I could see the color green. It didn't change any other color, but it changed everything. I love NeXT very much and have a very strong drive to see it succeed. But it's like a switch flicked and I could see green for the first time.

Photographer Julie Chase filmed Steve Jobs with a 1/2" Beta video camera. Frames from the video were captured with a NeXTdimension board and Keith Ohlfs's NeXTtv software.