Birth of the PC Clone
by Marshall Burns
I had avoided computers throughout my college days at MIT because somehow I sensed they would suck me in and I would lose focus on the physics that I had decided to concentrate on. After finishing my degree I moved to Los Angeles and dabbled in this and that for a couple of years. Finally, it was time to get serious about my future and I decided that I would resume my physics education and go for a PhD. However, one small problem, I had no money. I looked at the world around me and decided that computers were, in the vernacular of the day, where its at. That was in July 1981, about three weeks before IBM made its Earth-shattering announcement of the Personal Computer. Talk about timing!
How do you start a career in a new field you know nothing about? I decided the answer was sales. Olympic Sales Company in Pasadena sold Apple computers alongside jewelry and board games. Even without any experience, they hired me to sell computers because they assumed that with a degree from MIT I could figure it out. They were right. When I made a sale to a nice couple on my first day, I asked if they would like me to set the computer up for them there in the store so they would know what to do when they got home. They accepted, so I openned the manual to page 1 and, reading out loud, started following the instructions.
When the PC was announced on August 12, there were three places to buy it: IBMs own retail Product Centers, Sears, and ComputerLand stores. There was a ComputerLand a couple of blocks down the street from Olympic Sales, and I started visiting to learn more about the famous new product. When shipments began in October, the back wall was stacked floor to ceiling with grey and white IBM boxes as the store struggled to keep up with demand. The following month, I took a job at ComputerLand and started selling those boxes myself.
For the next six months, I sold IBM PCs and Apple IIs side-by-side. While PCs flew off the shelves, the Apple competed well on price and I sold about equal numbers of the two. It was fascinating to see Apple benefit equally from the PCs launch because the IBM name drew people into the store who had never before thought of owning a computer. Once they gave the idea serious consideration, many of them decided to give it a try with the less sophisitcated and less costly Apple.
The PC sold at a pace that caught everyone by surprize, especially IBM. The stealth team that hatched the bold plan for a low-cost IBM computer had projected sales of 250,000 computers in its first five years. Within three years, the bulging factory in Boca Raton, Florida, was shipping that many PCs every month!1
The IBM PC combined the conservatism of its Big Blue heritage with the radical adventurism of the 1970s in which it was conceived. No component-level technology was invented to develop the product. Everything, from disk drives to the central processing chip to the operating system, was acquired from third-party suppliers. The idea of combining components like these into an inexpensive, mass manufactured package was not new, having been done by Apple, Texas Instruments, and many others. The electronic architecture of the computer was new, but offered no innovations to please engineering enthusiasts. Instead it was designed for the iron-clad reliability that was the hallmark of all IBM products before it.
As a ComputerLand salesman, my job was somewhat analogous to waitering in an upscale restaurant. Every IBM PC was made up by starting with a base system and adding items from a menu of components. Sometimes, when business was especially heavy, I would assemble systems myself in order to get them out the door. This is where I got the experience that allowed me to set up my own business a few months later.
There was a strange structure to the pricing. IBM offered a bare-bones computer for $1,265, called the cassette system because instead of a disk drive it provided a port for customers to hook up a tape recorder for data storage. IBM probably intended this system to compete on price with the Apple II, but it was a pretty sorry comparison. Almost no one bought the cassette system. Instead, they bought the computer with one or two disk drives. The retail price of a two-disk system with 64k of memory was $2,895. But in the back room, we assembled those computers starting with the same bare-bones cassette system, and adding a disk controller, cable, and two disk drives. Had the customers known enough to order these components separately and install them themselves, the total retail cost would be $2,225. Quite appart from the profit margin in the price of the components, the store was making $670, or a 30% additional mark-up for the 10 or 15 minutes it took to put the system together. Therein lay my opportunity.
The 30% markup available from assembling retail-priced components was only the beginning of the story. As mentioned above, the PC was made from standard, off-the-shelf components. Probably because it was not expecting to sell a lot of units, IBM didnt even bother to have the components relabled with its own name. If you opened up a PC back in those days, you would have seen the brand names Tandon on the disk drives and either AMD or TI on the memory chips. Thus it was feasible to obtain some of the components directly from their manufacturers at wholesale and to assemble a computer that was identical in every way to a PC you would purchase from an IBM Product Center, Sears, or ComputerLand.
That was the basis of the business that began as Marshall Burns Computer Sales in Pasadena, California. On the whole, I found that by purchasing proprietary components directly from IBM at retail and assembling systems from them populated with third-party components identical to those used by IBM, I had about a 40% margin between my cost and IBMs advertised price for that same configuration. In my price structure, I basically split that diference with the customer, offering my systems at about 20% below retail, leaving a 20% gross profit in my sales.
While I had heard of one other person selling systems on this basis, as far as I know I was the first to advertise them. My ads appeared first in the LA Times and then in the Wall Street Journal. Orders came in from all over the country. I became known as an unauthorized dealer or grey marketer because supposedly only IBM, Sears, and ComputerLand had the right to sell the IBM PC. But the fact is, those outlets were the only ones that were allowed to buy their components from IBM at wholesale. Although IBM put pressure on ComputerLand franchisees not to sell the proprietary components to me at a volume discount, the IBM Product Centers never refused to sell me large quantities of those components directly at full retail.
At this point, the story might have gone differently had I spent some time around the Sloan Business School when I was at MIT. But I was pretty much a gear-head and had my heart set on going to graduate school in physics. I was excited to get accepted by the University of Texas at Austin, where John Wheeler was in residence, the Einstein protege who had coined the term black hole. So I hired a young fellow to learn the ropes of the business and moved him to Austin to run it for me. We incorporated there as Ennex Technology Marketing, Inc., got some office space, and he hired a few sales people to start growing the business. Things went pretty well for a while until he apparently decided that he was more important to the business than I was. Since I was spending something like 80 hourse a week trying to get back into the groove of physics, he was probably right. But when his superior attitude turned to embezzlement, it was time to shut the thing down.2
Eventually, other people started to get the idea of what I had been doing. One of them was a freshman at UT Austin who started PCs Unlimited in his dorm room. His name was Michael Dell.
Although I stuck with my plans in physics and left the fortune building to Dell, I hadnt done too badly. Starting with about $1,500 in the bank, the business had operated for only about eight months, during which time, we sold 99 computers for a total of $298,000. Our quality record was perfect, since we did a 24-hour burn-in of every system before shipping and not one computer was returned defective. When it was all over in December 1982, I had a small condominium on the East Side of Austin and enough money to meet my July 81 objective of financing my graduate school education.