Date: 30 Aug 85
To: List
From: John Day
Copies:
Subject: Poor Murray was just
a Product of his Environment

Recently, someone had a rubber stamp made for John Pugh that said "It's Murray's Fault." And to be sure much of the mess, the squandered money, the weak product line and the wasted two years can be laid at Murray's feet. But then is it really all that surprising that Murray had the problems he had.

It has always seemed somewhat incongruous that Codex was being lead in a new entreprenurial direction by someone from IBM. Since when does IBM know anything about being an entrepreneur! Not since about 1950.

Today, IBM's ability to make money is predicated on the fact that they have 75% of the market, not on their ability to act entrepreneurially. In fact, IBM's strategy is virtually the opposite of the approach an entrepreneur would take. To use Day's ecological definition of an entrepreneur: An entrepreneur is someone who recognizes an unexploited niche and exploits the hell out of it. Consider all of the unexploited niches IBM has not pioneered: big mainframes, Univac; minicomputers, DEC; timesharing, everyone else but; PCs, not them; networks, Telenet, and others; etc. IBM always waited until someone else had pioneered the market and then moved in as the heavy and took a piece, albeit a big piece.

The IBM strategy won't work for the rest of us.

In all fairness, IBM was the first, and practically only company, that recognized very early that for the first twenty or thirty years in this industry (until a generation of upper managers raised on the new technology were in place), the guy who would sign the bottom line for several million dollars wouldn't know beans about what he was buying. As long as the sale was made and customer service could keep the stuff running at an acceptable level and the pain of changing was perceived to be great, it didn't matter whether it was well-engineered, efficient, or any other attributes that are admired by the Protestant work ethic. IBMs approach has been to get a solid piece of the market while the buyer was ignorant and lock him in, before he was sufficiently sophisticated to understand what he was buying.

As everyone knows IBM does two things very well, sales and customer service , not engineering. If one wants to recruit from IBM, these are two areas that are worth looking at. Given the stress on these two areas, it is clear that there are no selective pressures (in the sense of natural selection) on the engineering staff. Sales sells without stressing the technology only that IBM will solve your problem and customer service makes it work regardless of what engineering did. There is no pressure on engineering to produce a good product. The only selection is in playing the organization game, not in producing product.

Two groups well-known for their intellectual prowess and deep insights, MBAs and trade press journalists, seem to have convinced everyone in the industry that IBM and by association the people that work for them have a corner on how to make money in this industry. There have been any number of doomsayers recently saying that IBM has the industry wrapped up and the rest of us may as well pack up and go home. However, it is worth noting that the computer industry is not a single market anymore. There used to be only mainframes, but now there are minis, micros, communications, PCs, factory automation, office automation, etc. IBM does not have 75% of all of these markets. In fact to go back to a previous point, IBM is most successful in those markets where the buyer is the least sophisticated. Consider the IBM PC, it is old hardware with a 25 year old software architecture.

It is possible to emulate some aspects of IBM's success. Two approaches have been tried: slavishly follow them, and compatible products that greatly out perform the IBM equivalent. The first hasn't been very successful, consider RCA and others. The second only works as long as IBM doesn't consider the market significant. Once it is signficant, IBM reasserts their position, consider Amdahl and various disk manufacturers.

The way that seems to work is to develop your own story (which must include how it deals with IBM) that is consistent and addresses customer's problems, sell it well, and support it well. As the field matures, it becomes more and more unlikely that anyone (even IBM) will be able to occupy 75% of any market. IBM's sheer size ensures it a healthy piece, regardless of what they do. And as the sophistication of the buyer increases, IBM will have to rely more on their locked in customer base and in changes in their strategy.

One must keep in mind that these are not absolutes. Every population is a Gaussian distribution. There are "plus one sigma events" in every population. But even they need time to adjust to the new environment. About a year before he left, in a moment of candor that was most perplexing, Mike Doss said to me, "Most IBMers fail in their first job outside IBM." Poor Murray just had problems adjusting.