Virulent Word of Mouse

October 5, 2003

Book Review: Jumping on Bandwagons

Filed under: Book reviews — Shane Greenstein @ 7:46 pm

Think back to the days when the world’s largest telephone company, AT&T, was also the world’s largest regulated monopolist. It was an insanely profitable company, generating billions of dollars in revenue a year.This same company organized one of the seminal commercial failures of the 1970s, the Picturephone. AT&T had followed all the rules, read all the technological forecasts, and-so its managers thought-began to develop the product of the future.Yet, the Picturephone had one huge problem: Nobody wanted to buy it.

The whole affair was a spectacular and embarrassing commercial disaster.To be sure, there have been many other failures of new ventures in communications since then, but none quite as embarrassing (with the possible exception of the recent disaster at Iridium, the satellite phone consortium organized by Motorola).

Back in those days, AT&T put some of its revenue into the world’s largest privately financed laboratory, Bell Labs.Those labs employed economist Jeffrey Rohlfs. Motivated by the events surrounding the Picturephone, he published a paper in 1973. At the time, few noticed it. But it has grown into one of the classic academic papers about network economics.

Three decades later he has written a book summarizing the lessons he has learned; the book is called Bandwagon Effects in High Technology Industries  (MIT Press, 2001). Few writers hit the sweet spot between accessible writing and applied frameworks. Rohlfs has found this balance. He explains the economics of bandwagons clearly, succinctly, and smoothly. The book contains a unified framework, illustrative examples and extensive discussion. Moreover, these different types of presentations contribute to the reader’s understanding of each topic.

Catalytic event

Rohlfs’ inquiries began in the 1970s, and his book begins with this biographical tale. Ever the economist, Rohlfs was fascinated by the decision to adopt or not adopt the Picturephone. He focused on an asymmetry in the value proposition for a user. Value to an individual user from a communications device was trivial when the network was small, but much larger when others used the network. Explaining this example will illustrate his general approach.

Three pieces of economics emerged. First, some users define the asymmetry in the value proposition. Call this a critical mass. Once a network’s size becomes greater than this critical mass, a network’s value proposition changes dramatically. Below this critical mass, many potential users do not find the value large enough to motivate a purchase. Above the critical mass, they do.

Second, AT&T was not going to invest heavily in widespread infrastructure to reduce the costs of providing the network until it was certain that the network would be large. Of course, nobody was going to adopt the Picturephone until AT&T built the infrastructure. It was much easier to rely on the voice telephone, whose infrastructure was already in place. So it goes: Networks are hard to start, but, paradoxically, they are also hard to stop once started. That is, it is hard to build a network as long as its size remains below a crucial critical mass, and it is hard to stop an established network from growing once it surpasses this crucial size.

Third, multiple outcomes are possible, so the sequence of adoption decisions determines which option emerges. Rohlfs considered several of these potential scenarios. For example, if some risk takers purchase when the network is small, this might motivate risk-averse users to follow, starting the whole process towards building. Alternatively, other big movers could play a key role. The Department of Defense (DOD) could use the Picture- phone for its own purposes, and as a by- product, motivate the building of support infrastructure. Experience gained from supporting the DOD’s needs could bring down the costs of a big network, again driving up adoptions by consumers. Indeed, this is precisely how the Internet started.

The intuition in any such example is compelling: Building a successful network involves starting a bandwagon so the net work achieves critical mass. Bad engineering is not necessarily the cause of commercial failure; the commercialization strategy also should bear responsibility. It must account for the role of critical mass in shaping the value proposition.

Examining one industry after another

The book’s inquiries provide in-depth description and analysis of several other examples that display a similar phenomenon. Rohlfs pays particular attention to the use and development of the fax, television, VCR, CD, PC, and Internet. Most of these stories are familiar, but I cannot think of any book that puts them all together in one place with this much depth.

Rohlfs finds a balance between his framework and the details of each case. He wants readers to appreciate both the similarities among situations and the details of disparate markets. For my taste, the success of this book lies in this balance. Rohlfs goes deep enough to show the reader where the framework succeeds and where it has difficulty. Yet, he is succinct enough about the details without being superficially glib.

I am particularly fond of his analysis of the fax machine, partly because it embodies a few key lessons that are largely underappreciated by MBAs. Rohlfs talks about the multiple starts and stops that preceded what eventually became the successful product. Then he goes on to describe how the fax achieved critical mass in modern economies.

Proponents of something like a facsimile machine knew what they wanted it to do for quite a long time, well before the successful design emerged. They simply did not know how it would be done. The process of designing a valuable system required market-oriented trialand- error. It involved interplay between engineering and learning by using; in other words, what did experienced users actually find valuable and how should engineers redesign and enhance those features?

The bandwagon did not develop smoothly. Many large corporations adopted facsimile machines for their own use to fax from one building to another or, eventually, between floors on a skyscraper. These early adopters did not concern themselves with the formation of critical mass in the country as a whole. Popular usage only emerged later and in a somewhat uncoordinated fashion, after many large companies had adopted the fax for their own use.

The eventual successful design found a balance between perfect engineering and a cheap, quick-and-dirty design. Critical mass emerged because users did not have to adopt an entire system, just one inexpensive piece at a time. Moreover, the national critical mass was the compilation of many smaller critical masses, each developing for its own purpose.

The analyses of TV, PC, CD, VCR, and Internet all contain the same balance between detail and analytical rigor. All are nicely done.

Intellectual struggle

An intellectual struggle lies below this book’s surface. To his credit, Rohlfs has a problem that most big thinkers could only dream of having.

In short, Rohlfs opened a Pandora’s box with his first paper in 1973, unleashing a torrent of different insights. Marketing books today spend multiple chapters describing tipping points. Standard undergraduate microeconomic textbooks today include a chapter on network externalities. There are many grand economics theories in which increasing returns plays a crucial role. This topic even trickled into antitrust law, as prosecutors and Microsoft lawyers debated whether Microsoft acted legally when it built critical mass for Internet Explorer and denied it to Netscape Navigator.

Against this background, Rohlfs’ book claims a bit of intellectual paternity. Sensibly, he is not trying to be responsible for the actions of his descendents. Instead, he is trying to establish a modest claim with his academic brethren, as well as with a few national business reporters who have not done their homework. To his credit, he deserves paternity for examining this topic early, for returning to it often, and for providing a coherent and lucid framework for a disparate set of important examples.

Consistent with his career goals today as a strategic consultant, Rohlfs’ second goal is to propose a particular vocabulary for analyzing events. Again, to his credit, he provides a refined, yet powerful, vocabulary for framing and constructing applied analyses about bandwagons. His analyses are not doctrinaire, and each illustration is extensive enough to consider a variety of arguments.

To be sure, Rohlfs is most comfortable analyzing markets that are mostly done growing, not how to predict a marketing strategy for a nascent product of unknown future value. He is also up front about this difference.

In other words, high-tech markets mature in a pervasive perfume of uncertainty about commercial prospects, disagreements about the future, and unpersuasive articulations of particular visions about the future. Value propositions do not emerge overnight. What clues does an observer have for understanding when nonadoption occurs because of the absence of critical mass or because of a poorly designed product? Practitioners never find these questions easy to answer. Rohlfs is not sure he has the answer either, but he is convinced that understanding previous examples will help with those in the future.

I should also add that this book contains an excellent set of cases for the classroom. It provides a clean and clear benchmark for readers who do not want to wade through mathematical theories about bandwagons, but who want to understand the deeper issues. These are notable achievements for such an accessible book.

Published PDF version of this entry

Leave a Comment »

No comments yet.

RSS feed for comments on this post. TrackBack URI

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Blog at

%d bloggers like this: