9

Lately I've been curious about early personal computers and how they became popular. Something I've realized is that my idea of which computers were most prevalent, and when, was quite wrong. I grew up using MS-DOS machines in the late 80s and early 90s, and so I imagined the ubiquity of "PC-compatibles" went back farther than it seems was really the case.

So now I'm wondering if anyone has ever seen anything like market research about how many households were using which lines of computer (Atari/Apple/Commodore/RadioShack/what-have-you) in given years in the 70s and 80s? I have been frustrated trying to find anything like this. I have seen some sales figures (Jeremy Reimer's impressive Ars Technica articles seem to be kind of authoritative here), but that kind of information doesn't quite paint the picture I'm interested in. With Internet Archive hosting so many early PC user magazines now I wonder if there might be consumer surveys or the like buried somewhere in some of them, but I thought it might be worth asking here in case anyone knows a nice go-to offhand :)

Any reply appreciated, thanks!

EDIT: Specifying my question per Raffzahn's suggestions (thank you!). Maybe any definition of a PC will be arbitrary but I'll say that here I'm interested in microprocessor-based, commercially available computer platforms which saw significant (not to say exclusive) adoption for use in homes. By "platform" I guess I mean a CPU plus some supporting standardized architecture enabling reliable use of common software. Generally I think the Altair 8800 is often touted as "the first" (again, a tricky designator but hopefully the particular arbitrary meaning I'm choosing gives enough of an idea of what I'm asking about at least) such machine, with the "1977 trinity" (Apple II, Commodore PET, Radio Shack TRS-80) following more prominently.

Really, numbers for "big names" among such platforms (Altair, the trinity, Atari 400/800, Commodore's many lines including Amiga, Atari ST, and eventually the IBM PC and Macintosh) alone would satisfy my curiosity I think. I started down this rabbit hole after waxing nostalgic with a friend about the computers we played games on growing up, then got to wondering what the "most typical" hardware in use would have looked like in given years. I hope this makes it clearer what I'm hoping to learn more about :)

SECOND EDIT for Raffzahn: Sorry, I forgot to link the articles I mentioned. https://arstechnica.com/features/2005/12/total-share/ for Reimer's article, and at https://jeremyreimer.com/rockets-item.lsp?p=137 he offers a link to the dataset he collected and drew from for that article.

EDIT to answer Greenonline: Say (per Reimer's data) ~5 million IBM PCs/compatibles were sold in 1986, the first year in which that line accounted for a majority market share. Presumably that doesn't mean that in 1986, more than half of people who had a computer in their home were using an IBM PC or PC-compatible, because plenty of computers were still in use that were purchased in 1985, '84, '83... does that make sense? Also, the figure of ~5 million isn't broken down by home vs. business use; lots of machines bought for business use are included there. It appears the Commodore 64 remains the best-selling single PC product of all time, and I'd imagine a higher proportion of C64s were bought for home use than business use, relative to IBM.

To Mark Morgan Lloyd's question: I didn't have that distinction in mind; by "MS-DOS machines" I meant only PC-compatibles :)

6
  • 3
    It might be helpful to specify what is considered a 'personal computer' in this context and what exact time frame it is about. Also linking the mentioned articles might be a good idea to establish what base this is set on.
    – Raffzahn
    Jul 16 at 18:35
  • 2
    Why would the sales figures not paint an accurate enough picture? What data/info are you looking for that the sales numbers wouldn't show? Do you mean: How many households had multiple computers, and in those which computer was used the most? Please edit your question to clarify (and add the details and links that Raffzahn mentioned) - don't add info in the comments.. Jul 16 at 18:52
  • 2
    If the question is more “computers we played games on” then, at least here in the UK, we had an all-formats top 40 sales chart that’d instead give some sense of which platforms were being used to play non-pirated games. The consoles don’t really start obscuring the computer numbers until the early ‘90s. Though it would be to answer a somewhat different question, so mentioned as a comment only.
    – Tommy
    Jul 16 at 21:58
  • 2
    Not wanting to pick nits, but when you say "grew up using MS-DOS machines" are you inclined towards that including PC-compatibles or are you explicitly trying to convey the idea of machines that ran MS-DOS but made no attempt to be PC-compatible like the Sirius/Victor range? "Back in the day" there was software that came in multiple variants: "PC-DOS" which assumed that there was a BIOS ane memory-mapped screen and "MS-DOS" for everything else: Turbo Pascal was notable in this regard. Jul 17 at 10:16
  • 3
    "Also, the figure of ~5 million isn't broken down by home vs. business use; lots of machines bought for business use are included there." - Do you want to know only about non-business use? except possibly in the US, the vast majority of 'IBM' PCs were used for business, even if at 'home' during the 80's. Jul 17 at 10:57

1 Answer 1

6

It's hard to come up with valid numbers from a time when most of the machines sold were not accounted as PC nor any of what we consider, when looking back, consider a major player. Apple may provide a nice example of how small the numbers we're talking about were. Despite being the first of 'the big three', the Apple II only sold (*1)

  • 1977 - less than 600 Units
  • 1978 - 7,600
  • 1979 - 35,000

The real success only came after that. By June 1983, exactly 6 years after the first sale, the millionth Apple II was produced. Which by later standards again isn't much.

In contrast, Altair sold over 5000 kits during 1975 ... then again, it can be debated how many of those were built into something that would resemble a PC at least as much as a C64 is a one (*2).

All of this is dwarfed by what was built and sold by companies way less visible when talking about that time, like Siemens, Olivetti, HP, etc., as they didn't sell them as PCs, but office systems, intelligent terminals or 'Calculators' - and more importantly almost exclusively sold for business purposes - let alone due to usually way higher prices.

Those early years were much less like imagined today.


But yes, I would think Mr. Reimer's numbers are a good start. So, let's use them for some Gedankenexperiment: Until 1980 his numbers, which seem to be US based, add up to ~3 million units. Now let's assume all of them were really sold to private users, not companies, schools, university or institutes (*3). At the same time, 1980, there were ~ 203 M residents in ~81 M households. That gives a meagre

  • 1.5% of US residents (1 in 68), or
  • 3.7% of US households (1 in 28)

owning a PC. Not much is it?


(I'll see if I can add other data points)


*1 - I did research that numbers some years ago for a museum project.

*2 - The basic Altair was a very raw CPU board with no I/O and 256 bytes of RAM.

*3 - Personally I'd set that way lower.

8
  • Siemens, Olivetti, HP - that list should surely be extended to include Wang even if you don't go so far as to want to add any more of the many office system mfrs of the time
    – davidbak
    Jul 17 at 0:54
  • @davidbak Oh, that list could be very long - except that the question for one focuses on home use, where Wang (and others like Datapoint or MAI) wasn't exactly a major player and second explicit asks micro processors, something tose companies only slowly embraced. But yeah, main reason, it would be a very long lis.
    – Raffzahn
    Jul 17 at 1:01
  • 1
    @davidbak Olivetti sold quite a lot of PCs in the US - not to mention their PCs under AT&T brand. Also, yeah, why not adding some Euro-centric here names. I also don't know when Wang started to use (standard) micro processors (for MAI and Datapoint I'm pretty sure that only happened during the second half of the 1980s) - same reason why I left out IBM (but included HP).
    – Raffzahn
    Jul 17 at 1:18
  • 1
    @davidbak Bit slice != micro processor. Dedicated word processor wasn't a bad idea. they did sell well all thruout the 1980s, way into the 1990s and maybe even early 2000s. After all, the PC as machine for everything on every work place is a rather new development - mostly due the Internet. Most jobs didn't (and still don't) need a PC but a fitting tool. Just think the Z80 based Amstrad PCW. Introduced as late as 1985 but selling worldwide over 9.5 million until 1998. So, yes, there was a market.
    – Raffzahn
    Jul 17 at 1:29
  • 1
    @davidbak Didn't Wang start out with dedicated word processors? I remember some machines with IBM Selectric as front end. Otherwise yes. A BASIC machine would have carved into their more expensive machines, at first and at the lower end. But it would have been the right way to save and even extend their business into the 1990s and beyond. Their BASIC was great and the software valuable. They opted for the same mistake of most successful companies - locking in their customers instead of widening the user base with new low end open devices.
    – Raffzahn
    Jul 17 at 1:42

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .