Intel's `paranoid' Andy Grove offers insights on helping your company survive the Web, Java, and more
Only the Paranoid Survive gives sound advice on how can you can avoid dangerous, corporate pitfalls
How do you know when your business is in deep trouble? More importantly, what can you do about it? Intel CEO Andy Grove answers these crucial questions in his new book, Only the Paranoid Survive: How to Exploit the Crisis Points That Challenge Every Company and Career. He offers a framework to analyze your company and your competitors and sniff out your weaknesses. We apply his methodology to changes in client/server computing as the Web and Java take center stage and consider Sun's transformation into a server company.
My career has gone to new places since March 1993, and so, I'm sure, has yours. This magazine has always stayed on top of the latest technologies, but perhaps your career path has been a progression from technical details to broader technology strategy and business issues, rather than, say, from C to C++ to Java programming. We would like to be a magazine that grows with you. That's the idea behind this change in focus for the column.
Bill's Bookshelf takes its cue from the column I wrote this past August in which I reviewed Geoffrey Moore's Inside the Tornado. In my new columns, I will review books that deal with technology and business strategies. My stylistic inspiration is my favorite Sunday brunch reading, the New York Times Book Review.
I am fascinated by some of the big-picture questions that we technologists keep in the back of our minds, such as: how can information technology bring about new business opportunities? How can it facilitate change in existing businesses? What are the implications when an entire industry becomes IT-based? How can technology companies maintain strong growth when there's a "paradigm shift" every 3 years?
Questions like these strongly relate to more down-to-earth questions like: will I still work at my company X years from now? What will the company be doing? Is my industry growing in its use of technology? What about technology vendor Y: should I buy its products? How stable is it?
All of these questions relate to the fundamental issue of change. Technology is about change -- or it is useless. Because we, as a society, are not used to dealing with constant, relentless change, much of today's technology and business literature grapples with the idea of change. Various book authors -- technology company CEOs, consultants, academics, industry veterans -- have different perspectives, which we can examine and try to validate. This sounds to me like a nice way to take a break, for a few hours each month, from the day-to-day concerns that envelop us all. I hope you agree.
As a bonus, you will be able to order the books reviewed here online, thanks to SunWorld's relationship with amazon.com, a "virtual bookstore" on the Web, with a virtual inventory of more than a million titles. (You'll see the amazon.com logo and hyperlinks at the bottom of this column). We are proud to be associated with this innovative Web site and to offer you the convenience of instantaneous ordering. However, don't think for a moment that this arrangement forces me to review all books positively! Now on to the first book.
The benefits of paranoia
Andrew Grove is the widely-known president and CEO of Intel. The title of his new book, actually his second, is Only the Paranoid Survive: How to Exploit the Crisis Points That Challenge Every Company and Career (Bantam Doubleday Dell). From this title and the general look of the cover, you might think that this book is yet another one of those silly sources of ego-gratification (not to mention publisher's revenue) for successful, megalomaniacal executives who have deluded themselves into thinking that they have wisdom for the ages.
Grove, on the other hand, is a part-time professor at Stanford University's business school. As he implies in this book, the fact that he chooses to spend his time teaching business students is significant. He clearly enjoys teaching, and he has made considerable effort to ensure that he has a lesson to teach. The tone of this book is a combination of business school lecture and candid, even-handed, personal reflection. Clearly, Grove's business has been phenomenally successful, but he does not spend much time gloating over it.
In Only the Paranoid Survive, Grove offers a useful framework for finding answers to the toughest questions about change: how do you "sniff the wind" so that you know when your growing business is about to go down the tubes? How can you act to prevent the decline before it's too late?
Grove spends much of the book answering the first question. He talks about a point in the growth curve of a business called a Strategic Inflection Point -- the spot at which growth changes to flatness and afterwards to decline. In mathematical terms, it's the point at which the second derivative of the growth curve goes from positive to zero. You need to determine when your business is at this point so that you can take the major steps that are necessary to prevent the inevitable decline.
You might think, as I did when I first read this, that this notion is completely trivial. And it's true that many modern management books, especially those by academics or long-time management consultants, treat today's rapid reversals of business fortunes as some sort of newfound personal revelation. But there are two aspects of Grove's model that aren't so obvious after all.
First, he is saying that you should start changing the way your company does business before the numbers start to decline, not after the decline has already begun. While it's not hard to determine when decline might be setting in (say, a few quarters in a row of decreasing revenue growth) it is hard to tell the difference between a temporary setback and the onset of real trouble. Furthermore -- as we'll see -- it's entirely possible that your business is in trouble despite nice-looking numbers.
So how do you tell when to change course? That's the second aspect of Grove's book that is worthwhile. If you are a senior executive, you have two tools that will help you determine if you're at a strategic inflection point. One is a variation on the kind of standard competitive analysis they teach at business schools. The other is the people in your company.
Analyze to death
Competitive analysis, the brainchild of Michael Porter of Harvard Business School, tells you that you have to look at several factors to determine how your business is doing. The first three are fairly obvious. Your customers: how loyal are they to you and how demanding? Your competitors: how fierce are they? Your suppliers: do you have a choice, or do the few available suppliers have you by the throat?
Then there are factors that are less obvious. Your potential competitors: are there companies that could, if they choose, compete with you successfully? A perfect example of this would have been Netscape about a year and a half ago. Microsoft was not a serious competitor in the Internet-software sweepstakes, but it certainly could have been. Now, of course, it is Netscape's worst nightmare come true.
The next factor is the possibility that some other type of business can emerge that delivers the same business benefits as your products or services, but does so in a different way. FedEx, for example, had a great business delivering documents overnight. Now, dirt-cheap fax modems and Internet e-mail accomplish the same thing, relegating FedEx to delivery of packages and "when it's gotta be hardcopy." The video-rental business will face a similar threat once broadband online service providers deliver video-on-demand to the home.
Grove adds a sixth factor that is important in technology-based businesses: your complementors -- businesses whose existence is symbiotic with yours. Intel and Microsoft are the most obvious complementors in the industry, as are Oracle (database server software) and Sun (server hardware). In today's standards-based, horizontal computing market, you need complementors, because without them, customers won't have complete solutions.
Grove's larger point is that your business could be at a strategic inflection point if one of these factors changes permanently and by an order of magnitude -- or as he calls it, a "10X" change. Sometimes it's easy to tell when a 10X change is happening. Netscape saw this the moment Bill Gates unveiled his Pearl Harbor Day Internet strategy. In the field of databases, Oracle and Sybase may be seeing it now with the launch of Informix Universal Server, the result of its acquisition of Illustra's object-relational database engine.
For another interesting example of 10X changes that can roil businesses, let's look at the history of graphics for presentations. Anyone who (like myself) does a lot of presentations for a living understands the dramatic drops in cost and lead time that technology has enabled over the past decade.
In the bad old days, you worked with your company's reprographics department (or an outside service bureau) to generate color slides. They delivered as overhead transparencies or 35mm slides - and probably with errors in them. Then came good presentation graphics software packages, like Harvard Graphics and Microsoft PowerPoint. These gave you complete control over the look and content of your presentations, but you were still at the mercy of the reprographics people, who took your floppy disk and came back with transparencies, but still took forever and charged a lot.
Nowadays, you can plug your laptop directly into a VGA/SVGA projector or an LCD panel used with a standard overhead projector. You can make changes to your presentation at the last minute, using superb, latest-generation graphics software, and the cost of displaying it is nothing except for the cost of that projector or LCD panel. These are outrageously expensive, well over $5,000. They are actually worth maybe $2,000 at most. How can this be? An SVGA projector is nothing more than a slide projector with an SVGA rendering device inside it. You can buy a top-quality slide projector for $500 and an LCD panel monitor probably costs around $1,000. Thus, it seems that projector makers like InFocus Systems are enjoying nice, fat margins at the expense of their customers.
Yet IBM made a move recently to introduce technology that would knock thousands off the cost of hardware for computer-generated presentations. A recent addition to its Thinkpad line of laptop computers includes an LCD monitor that detaches and becomes a transparent panel for use with an overhead projector. The cost is high for a laptop, but it's almost equivalent to buying an LCD panel and getting a top-quality laptop for free. This could be bad news for InFocus and other such companies, who must grapple with a "10X" force in the form of an unanticipated competitor.
If you can sniff out developments like these that threaten your business before they do any significant damage, you can take steps to prevent the damage. How do you do it? The key is in the title of the book: essentially, you need to be paranoid. Grove provides a piece of advice whose importance cannot be overstated: in a large company in a fast-moving industry, the last people to find out anything really important are the senior executives who run the company. The first ones to find out are those at the "periphery" of the organization: sales people in the field, who lose deals; engineers and marketing people, who must invent new products to counter those of the competition; customer service people, who deal with dissatisfied customers.
Senior executives must make special effort to listen to these folks' messages of doom and destruction, or they will be insulated from the truth and wonder what happened while their company goes down the drain. The worst situation is a company whose organizational structure is extremely rigid and hierarchical, and whose culture is such that bad news gets sugar-coated as it is passed up the chain of command. By the time the message reaches the top (if it ever does), it has lost all of its impact.
Think of IBM before Louis Gerstner took over: it was a culture of excruciatingly formal slide presentations (done expensively and time-consumingly by IBM's reprographics department!) in which all real information became dissolved into a series of inoffensive, information-impoverished bullet-points by the time it got to anyone important. Top executives could not possibly have done anything about the decline in mainframe business if all they heard was sycophancy from their subordinates.
Grove's message is that it's vitally important to listen to "Cassandras" from the front lines, even if their concerns might be provincial and their messages repetitive. It's up to you to synthesize a group of seemingly petty concerns ("We lost the deal because our e-mail package doesn't allow you to edit messages in the outbox" or "They hosed us because the PPP dialer won't work with COM3") into messages that are broadly meaningful ("We will lose unless we achieve feature parity with our top competitor").
To do this, you have to create a culture in which candid communication is encouraged and not punished. Since this is somewhat counterintuitive, you have to hammer this message home and be 100 percent consistent. If you're in a bad mood one day and you snap at the sales rep from northern Indiana who says that he didn't make his numbers this quarter because "the new product has more bugs than a beehive," the squeaky wheel will squeak all over the company, and all bets will be off. It will take you months or years to restore employees' candor.
So what happens if you start hearing bad news from the foxholes? Does this mean you are at a strategic inflection point and it's time for a corporate overhaul? Not necessarily. A senior executive is by no means qualified to make that decision on his or her own, or even in conclave with others on the executive committee. Grove says that the best way to determine whether drastic change is necessary is through vigorous debate, through all ranks of the company. Let the religious wars be fought, and let a consensus be reached that determines the next direction of the company. Debate encourages inclusiveness so that when the inevitable change does occur, a minimal number of people are alienated from the strategy.
The history of Sun Microsystems during the past few years illustrates a corporate process that would make Grove proud. Like many of you, I used my first Sun machine back when "client/server" only existed in research labs, the Intel-powered PC only ran DOS, and the Macintosh was just a toy. In other words, I was a programmer, and for me, Sun was to "high-powered workstation" as Kleenex was to "facial tissue."
When client/server computing became popular in the commercial arena in the early '90s, Sun reinvented itself as a server company virtually overnight. Nowadays, there are certainly factions within the company who are wedded to the SPARC/Solaris desktop paradigm, but the dominant thrust of the business is in server sales -- just ask any salesperson or system engineer who deals with customers. I remember when this reinvention happened. It was as if Sun woke up one morning and became a server company, and the whole thing seemed as natural as a caterpillar waking up one day as a butterfly.
Yet that transformation was relatively simple. Sun's servers were, more or less, simply larger versions of its traditional workstations. The whole point of SPARC technology and the Unix operating system was that they were scalable for use on the desktop as well as on the large, multi-user server. Now, Sun faces a transformation that is potentially far more convulsive.
First, Sun's competition has changed. Whereas its traditional competitors have been other large computer companies with scalable, Unix-based technology -- Hewlett-Packard, Digital Equipment, IBM, Silicon Graphics, and a few bit players -- now the competition is Microsoft with Windows NT running on large multiprocessor Pentium machines, DEC Alpha boxes, and eventually other powerful processors. In other words, Sun is experiencing a classic case of the potential competitor becoming a real one, exerting a "10X" influence on Sun's business. Some people at Sun call this the WiNTel threat.
As a Sun insider, I can see how the company is dealing with the WiNTel threat. It's exactly as Grove describes the necessary transformational behavior. There has been a tug-of-war between "corporate" (marketing, engineering, etc.) and the people at the periphery, mostly in the field sales organization. The sales force says, "The battle with NT is over. We lost. Now let's look to the future." At the same time, we have a CEO saying silly things about our only "real" competition being Silicon Graphics (as a pure Unix company, now that IBM and HP have Windows NT strategies).
There are staunch Unix dogmatists within the engineering staff who believe that Unix is the one, true "open system." On the other hand, we have a technology officer -- relatively new to Sun -- who has said "Unix is a proprietary system" in front of several hundred field engineers who cheered in agreement. And at a recent training session on the WiNTel threat, the top engineers and market strategists from SunSoft were practically burned at the stake by field engineers who concurred with the many industry analysts who have taken Sun to task for its lack of a Windows NT strategy.
Sun corporate eagerly encouraged all of these internal debates and is taking grief from the field quite seriously. Although it remains to be seen how Sun can counter the WiNTel threat in corporate, client/server computing, it's gratifying to see that Sun is doing the right kind of thing to survive the threat through change without falling apart.
The client/server meltdown
At the same time, the entire field of client/server computing is about to decline as the three-tier, Internet/intranet/Java paradigm takes over. The Internet -- what does Porter's competitive analysis, as augmented by Grove, have to say about it? To a maker of microprocessors like Intel, it's certainly a "10X" change in competition: both existing (Sun, Motorola) and potential (any maker of Network Computers, such as Oracle and Sun), since Internet software runs on any hardware. It's a "10X" change in competition for operating system makers like Sun and Apple for similar reasons.
To Microsoft, the new paradigm could also be a "10X" change due to a new source of equivalent product/service. Desktop computing platforms are currently operating systems, and Microsoft supplies the dominant one (Windows). In the new paradigm, the desktop computing platform is the "Webtop," as supplied by browser and Java Virtual Machine makers (Netscape, Sun).
Grove uses his modified competitive analysis to position the Internet in the last chapter of Only the Paranoid Survive. He concludes, unsurprisingly, that the Internet is a highly significant development that should cause his business to change in equally significant ways, though he's not quite sure how yet (or maybe he is, but isn't revealing any state secrets).
Unfortunately, the Internet chapter seems too much like an afterthought. His conclusions about the Internet are not as neat and clean as those he draws, earlier in the book, about the last major transition Intel made: until the late 1980s, Intel was primarily a memory company. Then, under unbearable threat from Japanese companies such as NEC and Fujitsu, it decided to switch its emphasis to microprocessors. The rest, as they say, is history, and as they also say, hindsight is 20/20. The Internet chapter must have been an editorial mandate. And it's amazing how easy today's word processors make it to insert several random instances of "as we will see in Chapter 9."
The other major shortcoming of this book is that it refers to the same experiences over and over. Grove may be one of the most successful and talented technology leaders, but, as an Intel executive since 1979, he simply doesn't have that wide a set of personal experiences from which to draw. Like any good business professor or consultant, he presents his model in neat, concise fashion. But unlike the latter, he can only say how his experiences fit the model. He cannot compare it to other experiences that fit or did not fit. Not even in pseudo-sciences like economics can one get away with a theory extrapolated from such a small set of data.
There are many books out that present similar models of change in technology-lubricated businesses. Grove's is not particularly more elegant or incisive than any others. His greatest strength, exhibited in this book, lies in the personal skills and assets he brings to bear on his job as CEO. Unless his writing utterly misrepresents his real-life behavior, he is a reflective thinker who lives by two important codes of executive conduct.
The first is that wholesale corporate change requires that executives be reeducated, without shame or arrogance. When Intel changed from being a memory to a microprocessor company, it required that a critical mass of executives become "software people." It does not suffice to hire a few engineers steeped in the new technology, or even to retrain existing technical staff, and let them figure things out. Similarly, Sun's metamorphosis from a client/server company to a Web/Java company requires that the top executives become familiar with Web technology and do so quickly and publicly.
The second bit of behavioral advice is that the way executives use their time can be very revealing, and therefore they must be vigilant about their schedules. They must allow enough time to listen to the voices of doom from the periphery. They must also be careful to spend time on issues that really matter during times of crisis. Grove notes with some asperity that the typical schedule of an executive at a company in crisis has too many items of the "reality avoidance" variety -- speaking at a regional employee recognition dinner, attending an industry conference, etc. The image communicates more than anything else.
That is actually the most basic and indisputable message of Only the Paranoid Survive, and it's the same message that any book on corporate change in today's fast-moving world imparts: continued growth requires fundamental change. Change requires leadership. Leadership requires concise, consistent, cogent communication. Grove's well-written book is as good a source as any for that enduring advice.
If you have technical problems with this magazine, contact firstname.lastname@example.org