jbass@polyslo.UUCP (John L Bass) (02/11/88)
>From: aburt@isis.UUCP (Andrew Burt) > You have still failed to back up your "off-the-cuff estimates. Given that > your original posting assumed 200 million Americans when in fact the figure, > as I pointed out before, is 240 million, I am hardly willing to believe your > back-of-the-envelope numbers unless you explain what they are -- in *detail*. IF there were black & white market size numbers it would be easy to give you the PROOF you think you need. Unfortunately I haven't seen any numbers which are specficly labeled "the offical PC 1988 market size". Lacking such I BOLDLY used the following estimate DETAILED in my first posting: "Consider the US has a total population of about 200 million people, of which, less than 60 million are working adults who may have some exposure to using a computer in the work place ... many of those access a mainframe with a terminal. Maybe 1 in 3 (about 20 million) actually has or needs access to a PC (on their desk or shared use). "The real number is probably much smaller -- look around you for the next week and take your own sample of how many people in the general population have/need access to a PC for any reason." The art of estimating is a skill that used to be taught to all using a slide rule. THINK about what the above MEANS --- DO A LITTLE CRITICAL thinking. The above estimate of one PC for 10 people (20 million PC's for 200 million people) is not outragously large (or small) ... to claim the number is either as little as 1 in 5 or as large as 1 in 20 WOULD take some justification! The 20% error in US population size estimate is in the noise given the other error probabilities! In sampling my aquaintances I get a number of something larger than 1 in 15 ... that may be an quirk of this local. I can not in my wildest dreams believe a number in 1988/89 of 1 in 5 or roughly 40 million units (or 48 million units by the revised estimate of population size). After subtracting out youth under 15, unemployed, housewives, and retired folks that would make the ratio in the working class about 1 in 3 ... totally absurd! Try the following: pick numbers, at RANDOM in the phone book until 20 people answer the following survey -- Ask them "do you CURRENTLY use a personal computer or have need to use a personal computer on a DAILY basis?". Factor the result to market size .... post your finding to the net. > Ok, let's make some small changes in your numbers (within the error range > for your numbers? I dunno, don't know anything about your numbers!) > > So assume a 25 million unit market, 3 year useful life, and current > shipment of .8 million units/month current shipping, then we are talking about > a drop from (.8 M/mon * 12 mon) 9.6 million units/year now to a replacement > market alone of (25M/3) 8.3 million, a decline of only 13.5%. Great ... lets look at the numbers .... a 3 year useful life means the product is discarded as scrap at the end of 3 years. THIS IS not happening like it did with the first generation Z80 systems!!! Most of the early (1981/82/83) IBM PC's and early clones are still in use with minor upgrades like add-on memory and a hard card. The cost of repair or replacement for motherboards and drives is a minor operating cost at the current time. The life of the electronics is 10 years plus. The current PC boom is largely driven (80% or more systems) as typewriter replacements. In any case the current MS/PC DOS 8088 systems will run Word Perfect keeping a typist happy for atleast another 4 years. This will make the average life atleast 5 years even if the machine is scrapped. The used PC market is becoming a significant factor in attracting new users/owners as the original owners move up to 286/386 machines and applications. I expect the real useful life number to be between 5 and 10 yrs. depending on where the industry goes in the next 5 years. This is VERY LIKE the automotive industry where current stats show the median life of a car at about 10 years. A new car is obsolete for the jet set in a year (new model comes out), first-owners keep a new car an average of about 3 years, a typical car has 3 or 4 owners in it's lifetime. A typical person buys ONE new car in their lifetime. A typical new drivers first car is a USED car. So assume a 25 million unit market, 5 year useful life, and current shipment of .8 million units/month current shipping, then we are talking about a drop from (.8 M/mon * 12 mon) 9.6 million units/year now to a replacement market alone of (25M/5) 5.0 million, a decline of about 50%. The .8 M/mon number is too low in reality since IBM is shipping about 0.25 M/mon and Apple/Atari/Comadore about 0.25 M/mon combined. There seem to be ATLEAST 3 clones (AT&T, Compaq, Tandy, SONY, and imports) for every real IBM box. This makes the number about .25 + .25 + .75 = 1.25 M/mon or 15 M/yr. This would make the decline about 67%. > Surely there > will be SOME market for first time buyers. I seriously doubt that the > roughly 110 million households in the US all have computers; some will not > want them, but who's to say? My personal experiences indicate even lower > income households purchase VCRs for example. What's to say PCs won't suddenly > interest them this way? Sure some will ... that is where I expect cheap used systems to go. Consider that VCR's staturated in the market at 37% of US households last Oct/Nov after 5 years of strong sales (made possible by the video rentals). This is a $225 item that has strong appeal to middle/lower class households with children where the cost of going out to the movies is $30 for the nite. A good return on the investment, even as an alternative to cable. As of 1988 a PC hasn't a fraction of the utility of a VCR for most households ... maybe some compelling application/software will be developed in the next few years which will drive a stronger home market. A PC with printer and hard drive (usable configuration) is also about four times the cost of a VCR. I don't expect 1 in 3 households to have a PC anytime in the next few years. > The more powerful the machines become the more > markets they open up. Who'd of thought of the publishing market for PCs > when Apple II's and TRS-80 model I's were the state of the art in micros, > for example? And this clearly ignores the business markets -- new businesses > start, old businesses find new uses for PCs. Sure things will continue to get better -- but how fast and how compelling will the upgrade/replacement cost be for the installed base of PC's? We are finally at a point where computers will start getting more expensive (after the initial market crash shock) since volumes will get lower and technology cost improvements are slowing down. (graph in posting # 2) > I have several problems with this graph... (a) it lacks units on the > Y axis; (b) it lacks sources for the data; (c) are you trying to say that > "#" (number new sales) will be ***0*** by 1991??? The graph was an off-the-cuff guesstimate specifically labeled "not to scale" to help people put the "dizzying growth and pell-mell embracing of PC technology"(from Larry Cambell's posting) in perspective with the effect of an adjustment to market equilibrium and a brief resurgence in the market as the first replacement wave hits. If you read the labels you would note that "#" IS NOT "(number new sales)" BUT RATHER "first time buyer sales" of new systems which I expect to go very close to zero. Most of the "new sales" will be replacement systems for previous owners/users of a PC. > > > 1) the crash in 1982 took the analysts/industry COMPLETELY off guard. > > (no one did accurate market size projections - myself included!!) > > So why are your numbers suddenly more accurate this time??? I didn't have any the first time - nobody I am aware of did!! Of course, some of us learn from our mistakes .... the rest of you shouldn't have to repeat them .... > > > 2) Companies only buy analyst reports that project a rose colored > > view of the industry for their prospective investors and bankers. > > Any well run company is not going to want rose-colored projections. Maybe > the companies you've worked with have only wanted the good news, but only > a fool would plan this way; and a company with billion$ in sales such > as IBM is not run by fools. We may disagree with them, but they aren't > stupid. Pretty strong statement ... but I guess mine was too ... the world is not perfect ... sometimes people only buy what they THINK they want to hear ... or are only given what someone else THINKS they should/want to hear .... > > In general, I think you lack sufficient data to make so bold a claim. > While I am no marketing whiz, my understanding of demand curves is that > they tend to slowly reach a peak, then decline slowly, except with fads, > which peak and decline rapidly. I seriously doubt PCs are fads. I did not lack sufficient data to make the claim IT IS CLEAR to the experienced eye! You should study demand curves for the introduction of mass market products more carefully ... and see the errors in your analysis. PC's are not a fad ... just the right product, at the right cost, at the right time ... Just as VCR's and other trendy/fad high tech product introductions. > You are also ignoring the effect a drop in unit price has on demand. > Your market of "20 million" may grow to "30 million" if the price for the > same functionality drops 25% (hypothetical numbers). Your figures ignore this. If transit busses were dropped to $500 ea I doubt they would replace cars ... they are too expensive to maintain and cost to much in fuel to operate. Cats are virtually free, not every household has one. Old newspapers could be worth money ... yet people dump them in the trash every day. Schools turn down donations of IBM 360/50's because they are to expensive to operate and maintain. Even PC's require a need to aquire, otherwise they become just so much clutter among one's personal effects. The day that a personal computer was a valuable property is past, or very nearly so. In 1968 I would have given anything for an electronic calculator with a single "memory". Today I have a box of them that haven't been used in 5 years -- for good reason. In 1972 I would have given anything for a computer that would run basic ... today you can get a sinclar for free (or less than $5) at a garage sale -- who cares any more for a basic computer???. The production costs of computers are starting to reach bottom. The false econmomy of falling prices in this industry may revert to normal ... IE everything goes up by inflation each year ... certainly TV's, radios, and other electronics have started to. A sharp reduction in volume will affect this some. The Yen probably more ... I look at the market size in terms of need/want .... not cost ... If computers were free few people would have ten. > Again, I ask -- no, challenge! -- you to document your data and the algorithms > used in arriving at your figures. > If you want anyone to believe you then you must justify your assertions. > Otherwise why bother to post at all? Just for attention? I did the first time ... you were just too busy cutting at what you didn't want to hear to think critically about the statements made ... > I suspect you're ignoring many factors, but I won't even listen to your > claims until you give me some reason to. There are hundreds of factors I CHOOSE TO IGNORE ... they don't significantly affect the outcome, but do significantly affect the error probablility of the results --- and the ability to grasp the basics of the problem .... heck you couldn't even see the obvious in the population/market size estimates ..... why should I clutter up the problem even more???? Have fun ... John Bass DMS Design
steve@crcmar.crc.uucp (Steve Ardron) (02/18/88)
in article <1222@polyslo.UUCP>, jbass@polyslo.UUCP (John L Bass) says: > Xref: crcmar comp.misc:1249 misc.headlines:1571 misc.jobs.misc:682 talk.rumors:189 > Posted: Thu Feb 11 07:26:46 1988 > > IF there were black & white market size numbers it would be easy to > give you the PROOF you think you need. Unfortunately I haven't seen any > numbers which are specficly labeled "the offical PC 1988 market size". Lacking > such I BOLDLY used the following estimate DETAILED in my first posting: > > "The real number is probably much smaller -- look around you for the > next week and take your own sample of how many people in the general > population have/need access to a PC for any reason." > > The above estimate of one PC for 10 people (20 million PC's for 200 million > people) is not outragously large (or small) ... to claim the number is either > as little as 1 in 5 or as large as 1 in 20 WOULD take some justification! > The 20% error in US population size estimate is in the noise given the other > error probabilities! > > In sampling my aquaintances I get a number of something larger than > 1 in 15 ... that may be an quirk of this local. I can not in my wildest dreams > believe a number in 1988/89 of 1 in 5 or roughly 40 million units (or 48 million > units by the revised estimate of population size). After subtracting out > youth under 15, unemployed, housewives, and retired folks that would make > the ratio in the working class about 1 in 3 ... totally absurd! > > As of 1988 a PC hasn't a fraction of the utility of a VCR for > most households ... maybe some compelling application/software will be > developed in the next few years which will drive a stronger home market. > A PC with printer and hard drive (usable configuration) is also about > four times the cost of a VCR. I don't expect 1 in 3 households to have > a PC anytime in the next few years. > I also have a few problems with your projections. When I lived in the States, almost 100% of households that I knew had Home computers! Many, almost half, had more than one! As computers become more prominent, most middle class families buy >one computer for their children at school, many universities require students to buy their own computers, and many households don't resell their computers, they just get stuffed in a closet. You were also argueing against yourself later in your article, as you said that a change in price wouldn't affect the total demand, yet later you said that more people would buy a VCR than a PC because a PC system was 4X the cost. You also based almost all your arguements on using PCs as typewriters, an invalid assumption if I ever heard one. Many people use their machines for games at home, and many small businesses and not a few households use them (or will use them) for database applications, an application that is very dependant on the machine for how well it will run. With these applications in mind, a printer isn't needed, and not even a hard drive is required, bringing the price of the PC down to the level of a VCR. This is still rapidly falling with no sign of letting up. As was said by others, and I don't feel adequetly rebutted by you, there are still many other uses for PCs showing up. Many of these aren't brand new, and have allready completed much of their integration. Such things as computer shopping and banking have great potential for expanding the PC market when a PC and modem costs only a few hundred, along with the previously mentioned heating and light control, aswell as games, typing and databases that are allready there. I'd say these things made a PC a hell of alot more useful than a VCR that is only for entertainment. Stevie. P.S. please excuse the many typos and bad grammer, since I was in a hurry.
erict@flatline.UUCP (eric townsend) (02/19/88)
In article <1222@polyslo.UUCP>, jbass@polyslo.UUCP (John L Bass) writes: > > In sampling my aquaintances I get a number of something larger than > 1 in 15 ... that may be an quirk of this local. I can not in my wildest dreams I would imagine this is a quirk. The only even semi-valid use or your sample would be in making statements about the sample itself. There is really no predictive value in your sample. It sounds nit-picky, but it's rather important not to make hasty assumptions... > Try the following: pick numbers, at RANDOM in the phone book until > 20 people answer the following survey -- Ask them "do you CURRENTLY use > a personal computer or have need to use a personal computer on a DAILY basis?" This is better, but still not really valid each person in America does not have an equal chance of being chosen for the sample. A test of this sort using a list of names from the Census would be a little better... A phone book is better than "asking the guys at work", but still not perfect. > This is VERY LIKE the automotive industry where current stats > show the median life of a car at about 10 years. A new car is obsolete > for the jet set in a year (new model comes out), first-owners keep a > new car an average of about 3 years, a typical car has 3 or 4 owners in > it's lifetime. A typical person buys ONE new car in their lifetime. A typical > new drivers first car is a USED car. I think this is very valid with computers... I owned a C= 64, one of the first few thousand sold in America (how many of us payed $600+ for a C64?). Now I use a rather used 3b1. The C64 lasted me at least 7 or 8 years, I expect I'll use the 3b1 for 10 or 15.. If I don't sell it to someone else. As for how many households have computers, I would suggest to you the _Statistical_Abstract_of_the_United_States_ for the most recent year that they surveyed for computer ownership. -- Just say NO to skate harassment. | Just another journalist with too much If I wish really hard, will IBM go away forever? | computing power.. Girls play with toys. Real women skate. -- Powell Peralta ad J. Eric Townsend ->uunet!nuchat!flatline!erict smail:511Parker#2,Hstn,Tx,77007
roger_warren_tang@cup.portal.com (02/26/88)
"After subtracting out the under 15s, homemakers..etc" Flare lit tip off that Mr. Bass doesn't klnow what he's talking about. The civilian labor force ALREADY excludes these figures and is far above the paltry figure he posits.