GAITHERSBURG, MD (Nov. 19, 1999) -- "While selling 43,000 [Apple II] units was viewed as an IPO qualifying triumph in the smaller Apple of 1979, selling 140,000 Newtons was viewed as a failure in the giant Apple of 1994." -- The Innovator's Dilemma, p. 135.

I first heard about The Innovator's Dilemma, by Clayton Christenson, when its author shared the cover of the January 25 issue of Forbes with Intel (Nasdaq: INTC) Chairman Andrew Grove. The article explained that Intel originally created the Celeron processor because Andy Grove read this book. I got it. I read it. Now I'm telling everybody else who invests in, works for, or competes against a big corporation or a small start-up to read it. It's that good. It's one of those books that tells you things you didn't realize you already knew.

The Innovator's Dilemma is about how good management, by listening to customers and investing in what generates the greatest return, can ruin a successful company. It's also about how small companies turn garbage into gold. It explains how well-run and widely admired companies like Sears, DEC, and Harley Davidson can wind up on the wrong side of a David and Goliath struggle on their own home ground, and be vanquished at what seemed like the height of their strength.

It's happened hundreds of times. It happened when steamboats replaced sailing ships, when transistor radios supplanted vacuum tubes, when dirt bikes stole the market from heavier motorcycles, when hydraulic excavators displaced steam shovels, and it's happened repeatedly in the computer industry with the switch from mainframes to minicomputers to desktop PCs to laptops to handheld computers. It's happening right now as the book, music, and software publishing industries succumb to the Internet, and as the big steel mills cede their market to the minimills one niche at a time.

Each of these innovations is a "disruptive technology," a new way of doing things that initially doesn't meet the needs of existing customers. A "sustaining technology" produces an improved product customers are eager to buy, such as a faster car or larger hard drive, but a disruptive innovation produces something that existing customers can't use and don't want, which is initially WORSE than the old way for the original application. To sell the new product, the company must find new customers in new markets who value what the new technology can do. Some examples:

The early steamboats were expensive, slow, and unreliable. Companies sailing cargo across the oceans simply couldn't use them, even if they wanted to (which they didn't). So small companies took steamboats into rivers, where the ability to head straight upstream regardless of the direction (or absence) of the wind was valuable. After incubating the technology in rivers for a while, the steamboat companies expanded into lakes, which sailing ships didn't make much money in and thus didn't defend. Eventually, when steamboats did start hauling cargo and passengers across the oceans (where the REAL money was), they were faster and more reliable than sailing ships (even if they still did burn fuel), and the sailing ship companies were a decade or more behind the new technology and couldn't compete. The challengers swept the field.

Decades later, the sound quality produced by transistor amplifiers was (and still is) inferior to that of vacuum tubes. Amplifier companies knew that nobody would buy a tinny, buzzing transistor amplifier for their living room when vacuum tubes sounded so much better. So a tiny start-up company called "Sony" built portable, battery-powered transistor radios people could carry around with them. Sure, they sounded terrible, but who cared? Then, with the experience and revenue stream from the portables, Sony improved its technology to produce cheap, low-end transistor amplifiers that were "good enough" for home use, and used those revenues to improve the technology further and produce better radios. Today, vacuum tubes are only used in ultra-high-end amplifiers in concerts or recording studios. Vacuum tubes still sound slightly better, but most people find the much cheaper product adequate.

The new markets created by disruptive technologies start small, usually too small to interest a large company that needs to find billions of dollars worth of new revenue every year to keep growing. For a small start-up company, a $10 million contract is a huge win. For a multinational conglomerate, it's pocket change, and projecting that next year the market will triple to $30 million doesn't help much. Small companies view small opportunities as critical to their growth and success, while large companies view them as distractions from their core business. Disruptive technologies not only allow but encourage start-up companies to replace large, established ones because the size of the organization matches the size of the opportunity, and the small company can grow with the market simply by reinvesting their profits.

Even managers who understand that "this is the next big thing" (as Apple's John Sculley did with the Newton) simply can't force a small market to grow fast enough to satisfy the revenue needs of a big company. Trying to force a disruptive technology to meet the needs of existing customers immediately doesn't work either, as Chrysler found out by designing an electric minivan that cost $100,000 while the golf cart companies merrily sold their product to college students, shopping mall security forces, retirement communities, and vacation resorts. Yet, waiting for the market and the technology to grow and develop on their own cedes the lead to competitors, and by the time the big company tries to get in the game it's usually too late.

Disruptive technologies start by doing things for new customers, to whom the previous technology didn't serve. Often, several potential uses for the disruptive technology fail to attract customers before a workable niche can be found and developed. (Again, managers at a large company won't bet their careers when repeated failure is very likely, but successful entrepreneurs are used to dusting themselves off and trying again.) Then, like roaches under the floorboards, the new technology has a safe niche in which to develop where the currently dominant technology simply can't fit. (Portable vacuum-tube radios? A mom-and-pop store with a mainframe? Right.) Then the new company can improve its product and mount an attack on the low end of the existing company's product line, and repeat the process until there's no more low end left.

Well-run companies are vulnerable to this because they follow the money. The high-end is where they make their profits, so who cares if some cheap plastic replacement exists for a business they'd rather not be in? By attacking a business the entrenched competitor would rather not be in, and offering an attractive and lucrative avenue of retreat, the attacker can herd the larger company toward a narrower and narrower niche at the high end.

This is exactly what AMD was doing to Intel, and why the Celeron was necessary to stop it. When Windows 98 flopped in the marketplace and customers preferred to stick with Windows 95, Intel's continuing push toward faster chips created a market vacuum at the low end. Plenty of demand existed for slower and cheaper chips than Intel produced, and AMD stepped into that niche ready to use the steady revenue stream from low-end sales to finance a gradual upward attack. Intel put a stop to that with the Celeron, protecting its market at the expense of cannibalizing its own high-end sales. (Interestingly enough, Intel's own position was based on the rise of a disruptive technology: the microchip, a way of manufacturing electrical circuitry that couldn't initially compete with mainframe circuitry on performance but was much cheaper to produce in high volumes. And thus the rise of the PC.)

Steel mills similarly didn't want to make cheap concrete-reinforcing "rebar"; there was almost no money in it. Minimill technology produced very cheap steel by recycling scrap steel in electric arc furnaces, but the quality was terrible. The big steel companies couldn't use minimills for most of their product line, but for a few start-up companies minimills were good enough to make rebar. The big mills were almost happy not to be bothered selling a low-margin product with intense price competition and no customer loyalty, and they couldn't compete on price with the minimills anyway. They gave up rebar without a fight, and delivered a small but constant revenue stream to their competitor's research and development efforts.

As the minimills improved the quality of the steel they produced, they kept attacking upward, steeling (sorry, couldn't resist) the next least-profitable niche from the big steel mills. The margins of the big mills kept rising as they retreated into the high end, and their costs went down as they tightened their belts to compete on cost before giving up each low-end market. Paradoxically, with their market narrowing and their competition growing, the big mills posted record profits from their high-end focus and impressive but unsustainable growth. Anyone who knew the industry clearly saw what was coming, but investors didn't have a clue.

The same thing happened to Digital Equipment Corporation as desktop PCs gave spreadsheets and word processing to home offices and secretaries. DEC was busy doing its own upward attack, beefing up its VAX minicomputers to steal lucrative existing customers from IBM's mainframes at the high end. DEC had started by giving computers to midsized companies that couldn't afford a mainframe, but now their original customers didn't even need a minicomputer anymore. Perhaps a PC offered less bang for the buck than a minicomputer, but that was still plenty of power to calculate payroll for a small business that couldn't afford a minicomputer. The exact same weapon DEC had used to attack IBM was now back in the hands of IBM's PC division, and eventually DEC found itself a victim of the fate it planned for IBM. Ironically, DEC's remains were eventually purchased by Compaq, an example of dinosaurs mating.

Disruptive technologies not only allow but encourage start-up companies to replace large established ones. There's far more to the book than I could fit here, so I encourage you to read the Forbes article and The Innovator's Dilemma itself. If the Fool had a required reading section, these would definitely be on it.

Have a great weekend,

- Oak