Xref: utzoo comp.misc:1860 misc.headlines:2319 misc.jobs.misc:1060 talk.rumors:1028 Path: utzoo!mnetor!uunet!husc6!bbn!uwmcsd1!ig!agate!ucbvax!hplabs!sdcrdcf!csun!polyslo!jbass From: jbass@polyslo.UUCP (John L Bass) Newsgroups: comp.misc,misc.headlines,misc.jobs.misc,talk.rumors Subject: Doom and Gloom, as they say, revisited (computer market failures) Message-ID: <1177@polyslo.UUCP> Date: 7 Feb 88 00:39:53 GMT Organization: Cal Poly State Univ,CSC Dept,San Luis Obispo,CA 93407 Lines: 122 It seems that some people could not follow the train of thought in my first posting ... here it is a little slower ... The problem statement is basicly this: We have sold in the US something between 15 and 20 million PC's in the last few years. Based on some off-the-cuff population estimates I guess the US PC market installed base will stop about 20 million units. For various reasons this number might be as much as 25 million or slightly more ... I was hoping someone could make a sound case for a larger market size. At the current sales rate of about a million PC's per month it is simply a matter of months before new sales slow down. Another way to view the problem is: The replacement market equilibrium level for PC's is the market size divided by the useful life of the product. If we assume the US market size is about 20 million units for PC's, then the US market equilibium level is about 5 million units per year with a 4 year useful life. If we accept this market size of 20 million units, then shipments MUST fall from the current level of 10 million to 15 million PC's per year (50% decrease or more in unit volume, much larger decrease in $$$ volume). In both cases we are talking major crash ... 50% or more ... not just some major slowdown due to the general economy. If some bright person out there has a convincing argument that more than doubles the estimated market size I would like to hear it. For you visual folks, here is a simple picture: # first time buyer sales * replacement sales (3-5 yr delay of first time curve decaying to equilibrium) = market equilibrium level (market size divided by replacement interval) (not to scale) Volume ^ | | | # | | | # # | | # | * * | # * * * |=============#=====#==============*==#==============*= | # # * * * # -----#--#--*--*--*--*--*--*-----------------#--#--#--#-> TIME 78 80 82 84 86 88 90 92 94 The date line is my rough guess based on off-the-cuff estimates of shipments and market sizes. Because of the sharp increase in sales during '86 and '87 it doesn't really matter if my market size estimate is off by a factor of two -- we are shipping that many units a year. Thus the estimate of the crash date has about a 1 year window summer 88 to summer 89. Because of the sharp increase in sales during '86 and '87 most of the installed base will be less than 4 years old if the crash occurs in the next year. Thus relatively few replacement purchases will be made in 1988 and 1989. I got a lot of hate mail on the first posting, little of which addressed the issue properly ... most people just tried to delare that the crash could not happen because they did want it to ... best of luck to them ... As for the wise cracks about posting from an educational site, I recieved my BS in Computer Science from here a number of years ago and I've been taking a few classes reciently for the fun of it. I have been working in the industry for 15 years, working with UNIX since 1975, and did UNIX systems work in the industry at SRI International('76), ONYX('80) as a startup, and Fortune Systems('81) as a startup. Since 1984 I have consulted for a living with the DBA: DMS Design. The early UNIX record locking code started here at CalPoly as a semaphore/memory manager in the Lundy Display V6 UNIX driver, became locking at ONYX with a rewrite, and lockf during the /usr/group standards effort. My office system (dmsd) lost its disk during Dec, and not everything worked after reconfiguring the system and upgrading to latest releases of things ... postnews still doesn't work right. As for the less than interesting quotes on rose colored market projections consider two things ... 1) the crash in 1982 took the analysts/industry COMPLETELY off guard. (no one did accurate market size projections - myself included!!) and 2) Companies only buy analyst reports that project a rose colored view of the industry for their prospective investors and bankers. (doom and gloom doesn't sell well -- how do you present it in a POSITIVE way to banking, marketing and sales types??) plan ahead with open eyes .... Have Fun ... John L. Bass DMS Design (805)541-1575 Email polyslo!dmsd!bass The above is copyright 1988 by John L. Bass, and may be distributed freely without any change or modification provided this notice is included in all copies.