This computer market survey (“Computer Market Survey—Report No. 1”), authored by the editor of the journal in which it appears—Computers and Automation, May 1957—is one of the earliest of its type. It takes a peek into the future demands on the computer industry, looking at different industries, their growth potential their products and support, and the amount of money that they might be spending (“on products and services in the computer field in the next five years”). And what Edmund Berkeley saw in the next five years or so was--according to his survey--about $400,000,000 to $750,000,000 worth of computer sales.
The numbers come from a survey mailed to the subscribers of Computers and Automation: about 2200 were mailed out, and about 200 forms, or 9%, were returned. “The replies contain interesting, important and extraordinary information, showing the nature and size of the current market for computing and data processing machinery”, and it is upon these replies that the iterations of the report were established.
The categories that I'm interested in here are the types of organization, its size, and the possibility and computer/data processing purchases.
Some of the largest customers in the future for computers and data processing were reckoned to be:
Air transportation, ($2-4 million); electrical machinery ($1-6 million); "chemicals" ($10-15 million), "electrical apparatus and equipment" ($25-35 million); "gen(eral) mail order", $2.5-5 million); "naval reactors", ($5-8 million); "oil and chemicals", ($2.5-7 million); "maps", ($3-10 million); "petroleum, ($2-4 million); "research and atomic energy", ($5-7 million); "airplanes, missiles, reactors", ($35-50 million); logistics, ($4-6 million); "airplanes", ($7-10 million); "core memories, matrices, buffers, data processing", ($10-25 million); "airplanes, ($5-10 million); "computers", ($15-30 million)
The elephant in the small pink room stood to be "business machines", which came in at a whopping $100-200 million), which was about a quarter of the entire plausible expenses in the coming five years.
And some of the smallest:
"Naval ship construction", ($8,500-60,000); "rubber, plastics, etc.", ($100-200,000); "energy consulting services", ($80,000-120,000); "retail chain apparel, mens & boys", ($50-60,000); "basic agricultural) and industrial chemicals", ($180,000-200,000); "research & development", ($25-50,000); "data processing", ($100-300,000); "research & development" ($25-50,000); "aircraft components", ($10-40,000)
And "electronic research", ($2,500-$5,000)
Somewhere in the middle, but not yet discernible needs fell to:
"banking services", ($15,000-$2 million); "research", $25,000-$250,000); "custom design processes and systems", $1,000-500,000; life insurance, ($600-700,000); "department store", ($17,500-225,000)
And of course, "banking", ($15,000-$2 million).
This was all in 1957, fifty-five years ago. Perhaps a "computer year" is equal to something along the lines of a "dog year", like a 7:1 ratio, so perhaps when folks were trying to see ahead in 1957 in reference to the computer and computing needs it was sort of like trying to make a political prognostications for 2012 in the year 1662. It was just too much.
And that $450,000,000-$750,000,000 figure would be worth about seven times that, around $3-5 billion. Still a small figure, especially when you compare it to the defense budget for 1957, which was about $45 billion, which would make that figure just a few percent of the money spent on defense, and nowhere near what the amount would become. Still of course Berkeley was on the right path, its just that the potential for the use of the computer hadn't yet come to be realized.