The Sociology of Technology Adoption



by Howard Fosdick © 1992, 1999, 2006

This is the orginal article that introduced the concept of Hype Curves as published © Sept 1992 in Enterprise Systems Journal, subsequently republished © in Feb 1993 UniSphere and other tech magazines.

Introduction

Years ago I wrote an article for an obscure technical publication. The paper introduced a model for understanding how and why companies adopt new technologies. Called the hype curve, this model contrasts proven versus unproven technologies. It presents a simple methodology to evaluate how useful various technologies are to organizations.

The hype curve concept found widespread use. The Gartner Group adopted it as their paradigm for understanding technology evolution (often without crediting me). They developed the model further and renamed it "hype cycles." Silicon valley venture capitalists have employed it in evaluating and marketing technology. Groups such as the Tasmanian and Russian governments have used it for managing technological change.

Find more about its applications in the book Workflow Modeling: Tools for Process Improvement and Application Development, 2nd Ed. by Sharp and McDermott.

Here's the original unaltered article from the early 1990s...


To the long list of talents required of IS managers, add another: the ability to predict the future. That's right, soothsaying is a key attribute of successful IS managers. Spotting technology trends, knowing which will flourish, which will fail, and ultimately, which are applicable to one's IS department are critical to IS success.

Why this is so should readily be apparent. Buying into hardware or software technologies that do not become popular over the long term can lead an IS shop into a technological cul-de-sac. Poor purchasing decisions render a company's IS function dependent on proprietary or otherwise dead-ended technologies, with resulting costs to get back into the technological mainstream at a later date. Not understanding which technologies are likely to flourish and which will fail means investment in obsolete skills, hardware, and software. Only technologies with a future will have the "growth paths" that all vendors promise.

Example 1: In the early 1980s, a large manufacturing company presciently identified distributed computing as the wave of the future. Management realized they could capitalize on this trend to their business advantage. Unfortunately, their foresight failed when it came to how to achieve this vision -- they bought heavily into the IBM 8100 minicomputer. By the late 1980s, they faced an expensive conversion process to a technology with a brighter future: UNIX-based minis. The organization paid a high price as it switched to new hardware and new software, developed the skill set necessary to operating this new environment, and evolved the proper administrative procedures.

Example 2: Your company wants to move towards LAN-based client/server computing with a graphical user interface. You are assigned the task of choosing either OS/2 or Windows NT to make this happen. Which do you pick? A wrong decision could cost your company its first major initiative into LANs and GUIs. The consequences of your choice may also affect your career at the company.

The first step towards understanding how technologies will fare in the future is to understand how IS technologies evolve. Figure 1 postulates a curve for the progressive popularization, maturation, and decline of a technology that succeeds in mainstream IS. This curve is termed the technology's "usability curve," since it reflects the utility of the technology within IS departments.

Figure 1 presents a second curve that reflects the public attention given the technology. This "publicity velocity" or "publicity vector" is measured by such attributes as: the number of articles on the technology in the trade press, the frequency of conferences on the topic, how many "industry analysts" discuss the technology, to what degree vendor sales pieces and ads employ the technology's terminology, and similar measurements.

Hype Curves

Comparing a product's usability curve to its publicity velocity shows that many technologies are most discussed prior to their real use in IS. Ironically, once the technologies achieve full maturity, supporting perhaps hundreds or even thousands of shops on a daily basis, they receive much less attention in the media and other publicity forums.

This article cannot quantitatively verify figure 1. However, the histories of prominent software technologies over the past dozen years tend to substantiate these assertions.

Example 1: DB2. DB2 was announced in 1983, released in 1985, became usable with Version 1.3 in 1987, and had conquered the worldwide DBMS marketplace for MVS by 1990. Yet public attention was most heavily lavished on DB2 in the 1983 to 1985 timeframe. Why? Managers and technicians had to learn how to bring this new technology into their shops in a non-disruptive fashion. The "relational challenge" was a significant one. Now that DB2 is used at 6,000 shops for production applications, public attention has moved on to newer threats to the established IS order. DB2 is no longer deemed worthy of the press headlines it once received. Go to a conference like DatabaseWorld or DB Expo and you'll hear about client/server DBMS, object-oriented DBMS, and you-name-it. About everything, in short, but DB2. Quite an oddity for a product that owns its market. DB2's public profile has receded at the exact same time it has achieved market dominance.

Example 2: UNIX. UNIX was the darling of the press ten years ago. Several publications declared years in the early 1980s to be the "Year of UNIX" (one publication was so enthusiastic, it did so twice). Industry analysts and vendors trumpeted that the operating system would take over business computing by the late 1980s. (As is the custom in our industry, these prognosticators routinely neglected to mention their own monetary stake in the fulfillment of these predictions.) All this excitement occurred prior to any real viability of UNIX for business computing. About the only groups not promoting UNIX for their own selfish reasons were a loyal and sincere core of technologists, and IBM, which perceived its commercial interest threatened by UNIX.

By the late 1980s, UNIX had matured for business use. Security had been enhanced, products for administration, database management, and transaction processing were available, and the several features that had once made UNIX unsuitable for business computing were corrected.

As with DB2, UNIX's most intense publicity occurred at a time when the product was not ready for business use. But the publicity resulted in vendor enhancements that rendered UNIX eminently suitable for business computing. Once UNIX really became useful to business, and IS adopted it as a legitimate technology, the publicity died down. Meanwhile, many of us still wait for UNIX fulfill the IS market share claims its promoters predicted so loudly ten years ago.

Example 3: Expert Systems. A series of articles in Fortune magazine in the mid-1980s made no bones about it: expert systems technology would alter the face of commercial computing by 1990. Vendors eagerly followed this promising lead. Seemingly all software ads suddenly claimed embedded expert systems capability. And as always, industry analysts promoted this hype, stating that traditional programmers would be obsolete within a decade, their once highly-paid skills now worthless.

As the publicity died down, expert systems did penetrate business computing. But instead of displacing previous approaches to application development, they became one more useful tool in IS developers' toolkits. IS adopted expert systems shells in an evolutionary fashion, taking to the technology on a large scale only after shells integrated with the traditional tools of the IS environment. What finally lead to widespread use of expert systems was that they:

  • Were adapted to run on traditional IS platforms, like MVS and PC-DOS
  • Could invoke COBOL programs
  • Could be invoked by COBOL programs

Far from displacing traditional applications development, expert systems became a useful niche technology, making it possible for IS to address new kinds of applications. As IS successfully integrated this new technology into its traditional structure, expert systems started to yield concrete benefits, and public attention drifted elsewhere.

Why the Industry Misleads Us

Why are public claims for new technologies so out of kilter with actual usage? Why are technologies hyped out of all proportion prior to their actual usefulness? And why does attention decline once a technology really does become useful?

The above examples allude to the motives of various industry participants. The trade press, for example, needs attention-getting headlines to sell its wares. Reading about how object-oriented databases will put you out of a job certainly gets your attention, as opposed to useful information about commonly-employed relational technology, for example. But the press also fulfills a very legitimate need in helping IS professionals track new technology. This is of major benefit to IS managers, who need to understand new technology as it unfolds and judge its potential application to their shops. Unfortunately, somewhere in this process, a sense of perspective as to when the technology is ready for "prime time" gets lost. The amount of coverage on a technology is rarely proportional to its usage in real-world IS.

Industry analysts play a role in hyping technology. Being one myself, I can tell you that industry associations, trade groups, user groups and others will pay me good money to hear about the latest computing trends. But a presentation about mature technology, for example, about DB2 usage, even when it carries real value to thousands of users, is not financially viable. Flashy talks on trends are worth more than a good, solid technical presentation of immediate applicability!

Another factor influencing "industry analysts" is their own financial interest in vendors or technologies. Organize a technology conference, as I have, and you find quickly that speakers separate into two groups: those who speak for free (in order to sell a technology or product), and those who charge. Only the latter can be expected to be objective, yet they increase the cost of your conference considerably.

The presenting and writing that many "industry analysts" do is financially motivated. They often own stocks, take consulting fees, or even hold positions in companies dealing with the technology they hype. While in academic circles such behavior would be considered outrageous, in business computing such individuals, if successful, are admired for their savvy. This is fine, except for one fact: only the rarest of authors and presentors state up-front their financial loyalties. It is up to the unsuspecting IS audience to ferret out such sources of bias. Caveat emptor!

Vendors also participate in hyping new technologies. Obviously, the salability of their products depends on the perception of potential purchasers that the products embody the latest technology. Vendors thus abuse technological buzzwords, while working feverishly to bring some of the new technology into their products, either in legitimate or nominal ways. Both marketing campaigns and product development must conform to the current popularity of technology hype or vendors risk product failure. Many a good product has failed in the marketplace due to a lack of fit with current technological buzzwords.

Finally, technologists themselves oversell technologies, either willingly or unknowingly. We all know programmers who think whatever they are working on "will blow everything else away." While such enthusiasm is admirable in terms of productivity and professionalism, it skews objective judgement badly. Moreover, many technologists, by virtue of the detail orientation needed to work with technology and "make things work," lack an overall perspective and the breadth necessary to achieving it. So, technologists who are quite adept (in one narrow area), and who may be sought out for their technical judgement, in fact are poor persons to provide objective assessment of the greater trends in IS.

Finally, there are some technologists who are intentionally not objective. These persons buy into a technology for career or financial gain and use their technical authority to further the technology, instinctively believing that "one hand washes the other."

How Not to Get Misled – Technological Self-Defense

Awareness of industry dynamics in the promoting of new technologies, and the motives of many of the players, goes a long way towards ensuring your own ability to objectively judge technology's promise. You should no more assume that a press article, industry analysts' speech, or "world authority's" opinion is any more objective than that of your local used-car dealer unless you specifically know otherwise. Beyond this admonition, here are some other suggestions:

First, consider your own emotions and the they manner in which they may influence your judgement. Does the process of change, or not yet understanding a technology unconsciously bias you against it? We are all human, and as humans, our emotions sometimes skews our judgement. Identifying any possible emotions that may influence your judgement is the first step towards being able to discount their influence.

Second, consider the emotional reactions of your staff. The best DB2 programmer may not give you the most best assessment of OO-DBMS. The reasons might be programming enthusiasm, technical narrowness, job security, resistance to change, or a dozen other human factors. These factors do not render anyone's judgement worthless. But recognizing their presence helps to better factor this input into your own thought processes.

Third, seek references for companies that have actually (successfully) used any new technology you are considering bringing into your company. A technology that has passed the "hype phase" of figure 1 and entered the "early adaptors phase," will have some adherents who can point to their real-world experience and say "it works, because it worked for us." Ask vendors selling such technology for legitimate references you can verify. It is very much in their interest to produce them. If they cannot, you could be the unsuspecting beta test site for the product.

A good indication that a technology that passed the hype phase is if a not-for-profit users group exists for it. Any technology with a substantial base will likely have a users group you can attend. This is the best place to get "the real story," complete with triumphs and the problems others have faced to achieve them.

Conferences run by for-profit organizations fulfill a function more akin to that of the industry trade press: they help you identify and learn about technologies that may have promise for future IS use. However, they do NOT prove that these technologies are presently used in IS nor do they even indicate that these technologies will necessarily succeed in the IS marketplace over time.

Conclusion

Realizing that technology adoption is as much a sociological phenomenon as a technological one is key to identifying which technologies are appropriate for use in your shop. The "publicity velocity" a new technology attains is based on sociological factors: it is not a technical judgement. The self-interest of the industry trade press, industry analysts, vendors, and computer science researchers all intersect to create an intense publicity vector for a technology at a certain time in its life cycle. While ultimately any technology faces the test of its usefulness, this does not usually occur until sometime after the hype phase dies down, and real-world IS gains experience with the technology. Knowing when a technology has passed from the hype phase into real IS usage is essential to judging the viability of the technology for your installation.

As technologists, those of us in IS are sometimes prone to overly-simple technology assessments, ignoring the sociological context of our decision-making. Being aware of the sociological nature of the technology acceptance is crucial to understanding which technologies succeed in the marketplace, and why.

----------------------------------
[Added in 2023 --> Howard Fosdick is an independent consultant and industry analyst. Read his current articles here.]