MEGAMISTAKES: Forecasting and the Myth of Rapid Technological Change

by Steven P. Schnaars

Schnaars documents that there is no evidence to support the widely accepted hypothesis of accelerating change. He suggests that forecasters would do well to test their assumptions about the future with hard market questions grounded firmly in the present.

202 pages, Hardcover

First published January 1, 1989

Reviews

[Max · Rating 4 out of 5]

"When you get the urge to predict the future, better lie down until the feeling goes away."

Kinda repetitive at points, kinda suspect there's a selection bias towards the dumbest forecasts going on (it's funny to read, and serves Schnaars' thesis better). But overall it's a really useful exercise to reflect just how bad humans are at forecasting, both in general and in the domain covered in this book: predicting successful new tech.

The book kinda misses out on a Hansonian analysis: The public forecasts that the book analyzes are neither selected for how likely they will be correct, nor optimizing particularly strongly for it. (Schnaars evaluates forecasts predominantly selected from business magazines.) The 10-year forecasts seem to me more optimized for being inspirational, impressive, and surprising rather than their accuracy. Schnaars level of analysis would also not have picked out on whether there might be individual forecasters who outperform the mediocre performance of public forecasts.

Another angle I was wondering is whether from an investment perspective the forecasts actually balance hits and misses really well. Schnaars roughly finds that ~15% of technological forecasts turn out accurate. While I do think that Schnaars convincingly shows how many of the forecasts were predictably wrong, being correct 15% of the time might be good enough when choosing a portfolio that just needs a few big winners (e.g. successful forecasts included computers, the internet, photography, biological and chemical warfare, home education, mobile phones, commercial extraction of oil from shale, birth control, satellite communications).

The book is from 1989, so it's quite funny to read him making fun of technologies that later on became at least somewhat successful, such as induction ovens ("cooktop ranges"), boombox pillows, automated lawn mowers, air purifiers, Amazon.com, video telephony. He even lowkey made fun of a forecast of the fall of the Soviet Union, which is impressive timing.

Further notes

Limited success for forecasters:

"The most stunning conclusion to be drawn from a review of successful forecasts is that there are so few of them. Only about 20 percent of the forecasts examined could be classified as successes. Clearly, the record on growth market forecasting is not enviable."

"Mostly the successes have centered on a handful of innovations that went on to create huge growth markets. Generally, forecasters have anticipated the developments that led to the major growth markets of the past few decades. Forecasts for microprocessors, computers, VCRs, microwave ovens, and other [technologies]"

Reasons for Bad Technological Forecasts

Technological Wonder (Ch 2):

"The most prominent reason why technological forecasts have failed is that the people who made them have been seduced by technological wonder."

- Often exhibit a typical mind fallacy of consumers who don't care about new tech as much as forecasters do

The Zeitgeist Effect (Ch 5):

Forecasts from the same time period tend to be similar
- 1950s and 1960s: Super optimistic about economic growth, predicted widespread nuclear energy use (like cordless toasters fueled with nuclear power), jet engines were a big theme
- Supersonic flight was hugely overrated
- 1970s: Energy scarcity became a major theme
- Ultrasonics were predicted for everyday uses like dry cleaning and showers
- Space exploration was greatly overrated in the 60s

"In reaction to the 1957 forecasts, one broker stated: 'It isn't like today, where you have a feeling of helplessness, that problems can't be solved.' That reaction captures the malaise that Jimmy Carter talked about. The gross optimism of the late 1950s gave way to the pessimism of the 1970s, which in turn gave way to the renewed optimism of the 1980s."

"Unfortunately, to the dismay of construction technologists, the spirit of the times changed. The perception of nuclear energy in the 1950s and 1960s as a modern, too-inexpensive-to-meter, cutting-edge advancement changed to that of a dangerous, unmanageable technology in the 1980s."

Optimism Bias (Ch 4) - We want to believe:

"The excitement of extreme forecasts seems to serve another master. Such forecasts jar the imagination more than those for mundane innovations. Maybe we would all like to think that the future will be very different from the present. That way we might feel that we are really moving ahead."

And we underestimate the schlep necessary to implement even relatively successful new innovations.

"A review of past forecasts for video recorders and microwave ovens illustrates the length of time required for even the most successful innovations to diffuse through a mass market. It also refutes the argument that we live in times of ever faster change. Both products were introduced into commercial markets shortly after World War II. Both took more than twenty years to catch fire in a large market. The revolution was characterized more by a series of fits and starts than by a smooth unfolding pattern."

Examples of over-optimistic forecasts:

"Studies by industry groups have turned in similar performances. A late 1960s study by the Insurance Information Institute entitled 'A Report on Tomorrow,' published by National Underwriter, 'projects what life will be like in the 1980s.' It too foresaw automated highways, orbiting factories in space, undersea hotels, and ready-made houses delivered by helicopter. As the article on the study notes, the report 'takes a tone of cautious optimism.' Apparently, not cautious enough."

Videotex (a 1984 failed project for online banking and shopping):
- Made overly optimistic forecasts about user numbers and computer adoption
- Received investments from IBM but failed by 1988
- But looking at it from 2025, this is an example of how bearish predictions don't always work either:

"Forecasters are prone to mirages. Their burning thirst for trends leads them to see patterns when there are none. As a result they often see opportunities for spectacular growth without thinking the problem through. In regard to videotex systems, Andrew Pollack, writing in the New York Times, notes: 'People still prefer touching the merchandise in a department store to ordering by computer, and reading a newspaper to scanning a video display tube with their morning coffee.' He is using common sense."

Price-Performance Failures (Ch 6):
Forecasters fail to ask themselves basic questions:

"What additional benefit does this product offer over existing entries? Will consumers have to, and be willing to, pay extra for it? Does the product offer a benefit over existing products that justifies a higher price?"

- Often make optimistic assumptions about how prices will fall with increased market size

Methodological Problems in Forecasting:

Academics optimize for complicated mathematical modelling methods, that often make implicit assumptions that are wrong or are unable to capture commonsensical considerations

"There are no seers with special information or statistical knowledge in growth market forecasting. Contrary to expectations, forecasts based on a long and complex analysis are no better than simpler forecasts. It just does not work that way in growth market forecasting. Anyone with normal abilities who asks the right questions can reach the limits of accuracy. Growth market forecasting, by its very nature, is an egalitarian exercise."

Problem with Diffusion of Innovation research:
- Literature attempts to model past successful innovations with S-curve models
- Firms and forecasters think they can use an S-curve model for their innovation based on very few initial data points/sales
- This ignores the fact that most new products never grow past meager sales

False assumptions about innovation rates:

"Kahn echoed the often stated refrain that we live in a world that is changing rapidly. He cautioned his audience: 'All these predictions assume that present rates of innovation will continue. In a rapidly changing world, this is the only basis for predictions' (p. 113; emphasis added). It was not, and it was wrong. The go-go years of the 1960s quickly gave way to the recession-plagued 1970s, which themselves gave way to the economic growth of the mid-1980s. The dominant issues of the 1960s, 1970s, and 1980s were all very different."

Market Leader Myopia (Ch 9):
Criticizes market leaders for missing "obvious" innovations in their own sectors:

"In every instance, as well as in myriad others, market leaders were amazingly myopic in their perception of emerging markets in their own back-yards."

- Examples: Light beer not from Miller, ballpoint pen not from fountain pen makers, radial tires not from Goodyear, digital calculator not from manual calculator companies

Case Studies of Failed Forecasts

Transportation and Housing Forecasts:

"Probe" [1966] foresaw fantastic wonders ahead. Space travel played a prominent role in the TRW forecast. It predicted that the first manned lunar base would be established by 1977, followed by a permanent base in 1980. Also by 1980, a 500-kilowatt nuclear power plant would be operating on the moon! Getting there would be easy: By 1980 commercial passenger rockets would be operating. Finally, by 1983 a solar power plant would be placed in space, with wireless transmission back to earth."

"Housing would also change greatly. By the mid-1980s giant corporations would mass-produce low-cost, plastic (injection molded) modular housing. Those houses would contain many advanced features. They would have "real" air-conditioning, including germproofing. They would be not only fireproof but earthquake, tornado, and fallout-proof as well. Household chores would be automated. And, of course, the layout of the house would easily change with the family's needs."

Car design forecasts:

"Radical changes were in store for the auto of the future. Aerodynamics would increasingly be incorporated into car design—a clear, but rare, hit in hindsight. But windows would be sealed permanently, leaving air-conditioning as the sole source of ventilation. That is the way it was done on jets."

Critical of back cameras and navigation in cars:

"It is often hard to see the benefit offered by at least some of the high technology predicted as part of the car of the future. A recent version of the Cadillac Voyage, the sedan of the 1990s, had no rearview mirror. Instead, it came equipped with a backward-looking camera in the trunk and a TV screen in the dashboard. Another dash-mounted TV screen housed the inertial navigation system, which continuously plotted the car's position. A simple price-performance analysis would ask whether those innovations are superior to a silver-backed mirror mounted on a bracket and a paper map stored in the glove compartment (for which consumers still resist paying 75 cents), both of which are clearly cheaper."

Failed Technology Forecasts:
Entertainment technology failures:

"As one expert noted, 'The time has come when people will have the choice of watching Raiders of the Lost Ark, or being in Raiders of the Lost Ark.' Some were skeptical, but most manufacturers were afraid of missing out on the next 'moneyboat,' as one consultant called it. The boat never sailed. Mystery movies with different outcomes, like Clue, proved unsuccessful. More elaborate schemes never had a chance. The reasons were clear. Some interactive products required consumers to purchase hundreds of dollars' worth of equipment to interact with their TVs, including videodisks, which had failed badly a few years earlier. More important, research done at the time indicated key findings about the market the manufacturers intended to serve with these wondrous products: Viewers liked to be entertained passively and do not like to press buttons. In the end, viewers did not want more control over their TVs. They wanted to be entertained."

Picture phones were forecast as early as the 1960s:

"A study entitled 'A Long Look Ahead,' conducted for AT&T by the Institute for the Future in mid-1969, proved no exception. Big changes were in store for AT&T. 'The world of 1985,' the study warned, 'will be markedly different from today's.' There would be 3 million picture telephones in use in the United States, generating revenues of $5 billion. Other radical changes would occur. Telephone communications would increase because street violence would be so bad that citizens would fear traveling."

Forecasting Advice (if you insist on trying)

Advice for better forecasting:
- Avoid complicated models
- Simply write down what you know, what you expect and why
- Use broad brushes that make it easy to see and question assumptions
- Analogues are good (he hints at early versions of reference class forecasting)
- Use multiple approaches to reveal different assumptions and conflicts

Don't do probabilistic forecasting (!):
- It gives fake precision that is unwarranted
- You implicitly limit yourself to one scenario

He favors scenario analyses because they:
- Offer less fake precision
- Rely more on common sense
- Cover a broad range of outcomes
- (Imo this a fairly false dichotomy)

Recommendations for companies to react to the difficulty in forecasting success of new innovations:
- When developing new products, focus on a niche but certain customer base that likely would pay a higher price for the new product. Expand from there
- For established corporations, prepare for supplying new growth markets by investing in R&D and scaling capacities, without investing much in any individual product without learning from startups/competitors about product-market fit (IBM example)
- Invest in multiple products at once (e.g., RCA pursued both VCR and video discs due to uncertainty about customer choice)

[Jimmy Head · Rating 1 out of 5]

Irrelevant and myopic. Insults real futurist Alvin Toffler but doesn't have the courage to name the author of Future Shock and The Third Wave. Seen from only money manager view. Seems intentionally salmon swimming against the current. Anti-futurist, really?