The computer industry has seen ongoing technology evolution, both hardware and software, and this has led to a variety of different business strategies at different points in time.
The basic driving force for the computer industry has been the proliferation of application software that fulfills a real need.
1)The initial wave of this was with IBM and its 650/704/709 product line, along with Burroughs, Univac, GE, NCR and others in the US, and the initial wave of application software was for accounting (replacing paper records with electronic records, with the potential for far fewer errors, as well as a shorter time to see results).
2)This led to IBM developing its 360/370 line of so called mainframe computers, built with a common architecture and spanning several orders of magnitude in price/performance, and this was followed by Burroughs, Univac, NCR, CDC, and Honeywell/GE; the applications now covered engineering, manufacturing, sales and marketing, human resources et al, the entire gamut of activities in a modern business.
3)Minicomputers appeared to offer beer budget pricing for smaller businesses with champagne appetites, led by Digital Equipment Corporation, Hewlett Packard, Data General, Wang Laboratories, Prime Computer, followed by IBM et al.
4)Personal computers appeared led by Apple, Tandy, Commodore, and then legitimized by IBM, Dell, CompuAdd, AST, Sinclair et al.
5)Workstations appeared led by Sun MicroSystems, Apollo Computer, Silicon Graphics, Digital Equipment Corporation and IBM.
6)Cellular telephones appeared led by Motorola, Nokia, Apple and Samsung, et al.
Note that in each category new businesses appear, not simply the same names again and again. The Innovators Dilemma: When New Technologies Cause Great Firms to Fail, by Clayton Christensen, Harvard Business Review Press, 1997, suggested that the cause for this was that the dominant suppliers in one wave put their resources into supporting the existing product line and new technologies were not as relevant to the current business and did not get the needed attention; however, for startups, the new technologies were the basis for their business, and if they failed then the business failed, so the startups had every incentive to make new technologies a commercial success.
The underlying hardware technology advances particularly in integrated circuitry but also in manufacturing occurred roughly every seven years, as a new design point for computers became cost effective for a new set of applications: 1958 was the peak year for sales of 704/709 IBM computers, 1965 for IBM 360 computers, 1972 for minicomputers, 1979 for personal computers, 1986 for work stations, and 1993 for mobile phones.
ATT was faced with an anti-trust trial and elected to settle in the 1980 time frame. Two offers were studied: one was to keep all of the local operating telecom companies together along with the long distance company to make ATT solely a telecom service provider, and spinoff its engineering and manufacturing arms (Bell Labs and Western Electric). In hindsight this was the deal that should have been taken. Instead, ATT elected to spinoff the local operating telecom companies and to keep its long distance network plus its engineering and manufacturing so that it could enter the computer business. The settlement was announced in January 1982 and became effective on 1 January 1984. Bart Stuck was involved in setting the ATT Computer Strategy along with others, in particular Bruce Greenwald (now at Columbia University): the basis strategy was to launch a UNIX based work station/personal computer line, and to port UNIX to as many other hardware platforms as possible, which would encourage third parties to port their applications to UNIX or to develop new UNIX based applications. This was an idea that was ahead of its time: the bulk of the profits and value accrued to system vendors (IBM, DEC, HP, Wang, Apple) all of whom had proprietary environments, but this was undermined by the coming of the open systems IBM PC which had an Intel processor and Microsoft operating system, so that over time the value in computers shifted from system vendors to component vendors (Intel and Microsoft). However, with the advent of the mobile phone, Android and Apple IOS UNIX derivative system software became ascendant, while the servers used Linux and other UNIX derivatives.
In computing the question arises: does the data go to the computer, or does the computer go to the data? Historically this has moved back and forth: mainframe computers and minicomputers and personal computers typically had data come to the computer for processing. In 1964 DARPA funded Multics led by MIT, Bell Labs and GE, which attempted to provide a time sharing computer with terminals at remote locations from the computer. In the 1970s and 1980s commercial operations such as TymNet, CompuServe, and others offered time sharing services for sale. As computers proliferated the administrative operations became amenable to an economy of scale, whereby a small number of highly skilled engineers could manage the processing, storage, and communications functions of computing far more cost effectively than a large number of relatively unskilled individuals, and this led in the 2000s to the advent of cloud computing and the Internet.