Why the Strong May Be Late

19 February 2007 at 8:02 am Nicolai Foss 10 comments

| Nicolai Foss |

One of the favorite cases of technology strategy classes is the innovation of the CAT scanner. The first such commercially viable scanner was introduced in 1972 by EMI, by no means a major player in the diagnostic-imaging market. The EMI scanner was leapfrogged by General Electric in 1977. GE dominates the market today along with Siemens. EMI exited already in 1979.

Why was EMI the first to commercialize CAT scanners and not the more established players with the relevant core competencies? Typical explanations of this kind of pattern where strong players avoid pioneering a market involve “stupidity” (because of various biases the strong do not see it coming) or “lock-in” (for various reasons, the strong cannot respond effectively).

In a recent working paper, “The Timing of Strength,” Richard Makadok offers a third explanation, one that, he says, “is neither cognitive nor structural, but rather rational and game-theoretic: Compared to weaker firms, stronger firms can better afford to wait for the resolution of demand uncertainties” (the paper doesn’t seem to be online so you will have to mail Rich for a copy). In contrast, weaker firms cannot make effective comebacks after a delayed entry; hence, they have stronger motivation to enter early. Thus, at stake is a complex interaction of (commitment-)timing and strength (competitive advantage) that Rich nicely models with a variation of the basic Stackelberg model with different unit costs, the order entry being determined by a random draw, and uncertain market demand. This allows him to determine the circumstances under which it will be profitable for strong and weak firms respectively to enter a new market.

We have discussed Rich’s work earlier here on O&M (here and here), including Rich’s discussion of the various mechanisms that may produce profits. The present paper is clearly an outgrowth of the discussion in that paper, and perhaps a response to critiques that it is simply not manageable to present and examine the interaction between four different mechanisms of profit in a single paper. I am still struggling with the fairly involved math in this paper, but this is certainly a very worthwhile contribution. It is highly recommended for anyone with an interest in frontier research in strategic management.

A single (small) critical point: Teece’s seminal 1986 paper, “Profiting from Technological Innovation” should have been referenced. Teece also offers a “rational” (if perhaps not “game theoretic”) treatment, discusses timing strategies, invokes the CAT example, and develops an explanation of why the strong may be late that proceeds in terms of control of complementary assets (thus, he explains why the strong are strong).

Entry filed under: - Foss -, Management Theory, Strategic Management.

What Did the Legal Realists Believe, Really? Chip and Dan Heath on NPR

10 Comments Add your own

1. rtone | 19 February 2007 at 7:57 pm

Interesting, and certainly a consideration, but context and networking muddy the waters too much for mathematics! Great stuff for hindsight.
2. Richard Makadok | 19 February 2007 at 9:28 pm

Hi, Nicolai. Thanks for the helpful comments. Thanks also for the reference to the Teece paper — I had never read it before. It appears that he does make some points that are relevant to my argument. But as far as I can tell (and please correct me if you think I’m wrong about this — since there are so many different threads to his argument, it’s somewhat difficult to know for certain), ultimately Teece doesn’t seem to make the same argument that I do. He separates strengths into two categories — innovation and commercialization — and argues that early-stage success requires the former, while later-stage success requires the latter. But his argument still seems to be driven by the players’ abilities, rather than by their motivations (e.g., differential motivation to delay in order to gain the informational benefits of a “wait and see” approach).

Thanks again,
Rich
3. spostrel | 19 February 2007 at 10:39 pm

From Nikolai’s description, it sounds like the model asks IF you have a chance to go first, do you? That is the right theoretical question for medium-term strategy, because whether or not you get the chance to go first is exogenous in that time frame.

Some of the harder-core resource-based enthusiasts will insist on taking a long-term perspective and endogenizing the opportunity to go first–which, in the EMI case, means endogenizing who gets the inventive idea for the CAT scanner. It’s important to separate these questions so that the contribution of analysis like Rich’s can be understood.

If the case account is accurate, the historical, contingent reason why EMI got to move first should be congenial to Nikolai and Peter. It comes down to employing a brilliant individual, Dr. Hounsfield, who not only cooked up the idea and figured out how to implement the device but also, rather amazingly, invented a vastly superior second-generation machine when imitators caught up to his initial inspiration. There is nothing to suggest EMI had a superior organizatonal capability, just a superior employee. He won a Nobel Prize; I wonder what percentage of the profits to EMI he ended up capturing.
4. teppof | 20 February 2007 at 12:44 am

It seems that the Clay Christensen material also relates to the argument – the strong, incumbents, have dis-incentives to invest in new technologies as they fear cannibalization and want to appropriate as much as possible from past R&D (and, pace von Hippel and the customer-innovation stuff – customers demand incremental, not radical, improvements) – so entrants come in.

Steve: You bring up the right point – I think, as a first cut, that often there are underlying (or, at the very least, these should be accounted/controlled for) micro/individual-level reasons buried underneath what we call “organizational” advantage/innovation (we just rarely dig deep enough, empirically or theoretically) – thus I think we have a case here where reducing could help explicate (I think this is the case for much of strategy).
5. Richard Makadok | 20 February 2007 at 12:49 am

Actually, it’s a little-known fact (missing from the HBR case study on EMI) that the idea for CT was “cooked up” five years earlier (in 1967) by a team of researchers at Siemens, who studied it and then decided that, for technical reasons, it should be shelved for 10 to 20 years before attempting to commercialize it. This intentional choice by Siemens (one of the world’s leaders in diagnostic imaging equipment) NOT to pioneer the market makes my point about delay having relatively greater benefits for stronger firms.

— RJM
6. teppof | 20 February 2007 at 11:30 am

Hmm, I am always a bit sceptical about retrospective claims to inventions (Gore and the net), there is a team in Finland also that claimed to have invented the internet far before anyone conceived of connecting computers in that manner…
7. Richard Makadok | 20 February 2007 at 1:13 pm

It’s not actually a claim that Siemens invented anything. It’s just a claim that Siemens studied the idea and then rejected it.

And the source isn’t even from Siemens. The source is an article written by Walter Robb, who was head of GE Medical Systems — i.e., Siemens’ main competitor — from 1974 to 1986. It seems unlikely that he would have any motivation to exaggerate the technical accomplishments of his direct competitor. If anything, one might suspect the opposite — that he might have a motivation to downplay the accomplishments of his competitor.

In any case, the full reference is provided below…

Robb, W. L. 2003. Perspective on the First Ten Years of the CT Scanner Industry. Academic Radiology, 10(7): 756-760.

Thanks,
Rich
8. teppof | 20 February 2007 at 4:30 pm

Thats interesting – studied and rejected – that is an area that does not get looked at with regard to inventions, does it? (Some of the brainstorming stuff perhaps relates indirectly – but brainstorming, analyzing and selecting potential inventions seem like different capabilities). More generally, innovation indeed is extremely hard in terms of “imputation” (who created, how much, when, team production problems etc) and appropriation. Thanks for the reference.
9. Richard Makadok | 20 February 2007 at 8:22 pm

Good point… Maybe we could learn something useful about innovation by studying near-innovations — the same way that aviation safety researchers try to learn how to prevent accidents by studying near-misses.

— R
10. Nicolai Foss | 21 February 2007 at 7:27 am

Hi Rich, You are completely right that overall Teece is making a different argument from yours. I didn’t mean to imply that he was saying the same thing, only earlier; only that there are some parallels that warrants a reference to his paper. That being said II do, however, think (I don’t have access to the paper here, unfortunately) that Teece says in the paper (in connection with his discussion of timing strategies in the context of the product/process industry life-cycle) that control of complementary assets makes possible a “wait and see” strategy, so at least implicitly he would seem to invoke motivation and not just ability. (I may have picked this up from another Teece paper; there are a couple of published versions of the 1986 RP paper).