The incumbent’s (difficult) response to a disruption: Google and ChatGPT

Just a few weeks after the launch of ChatGPT, Google has launched its own solution, called Bard. Such a quick response from an incumbent threatened by a disruption is not a surprise. But does that mean that Google will be able to maintain its leadership position in its market? Not quite.

When considering the history of innovation, it is not surprising that Google responded so quickly to ChatGPT. The incumbent almost always responds to a disruption. In fact, it is rare that it is not aware of it. A notable exception is the mechanical cash register manufacturer NCR, which was completely taken by surprise in 1971 by the transistor. Kodak was not unaware of the digital disruption in photography. And for good reason: it was Kodak that had invented the first digital camera in 1975, and introduced products to the market as early as 1991. Nokia had a team working on smartphones and mastered touch screens. Airlines have almost all tried to respond to low cost.

Will Google’s response allow it to successfully maintain its leadership position in its market? To answer, consider Clayton Christensen’s theory of disruption, summarized in the expression Innovator’s Dilemma. As I mentioned in my previous article, the theory predicts that though there will be a response from the incumbent, it will be constrained by several factors. First, the fear that the response will compromise the incumbent’s legacy business; Second, that the response may pose a reputational problem (for Google, this would be the case if Bard started giving racist responses for example, as was the case with Microsoft’s solution in 2016); Third, the concern about the underperformance of the disruptive technology.

The issue of the initial performance of the disruptive technology

Typically, a disruptive technology tends to underperform the dominant technology, at least initially. The first Internet-based telephone communications in the late 1990s were of poor quality. A business customer could not use them. Rationally, the telecom operators of the time – whose business customers were the most profitable – rejected Internet telephony, and made way for new entrants like Skype. If it is underperforming on the main criterion, in this case quality, the disruptive technology nevertheless introduces new criteria where it is superior. For Internet telephony, it was obviously the price, since it was (almost) free. The question that then arises is to identify the users who accept mediocre quality because they value free service more. As we have seen, these were not the professionals. They could afford expensive communications, and quality was necessary for them. Those who were seduced by free calls at the price of mediocre quality were individuals, such as students far from their families. For them, it was Internet telephony, or no communication with their family. In other words, the quality, although mediocre, was sufficient for them (but not for professionals, who therefore refused it).

This shows something quite fundamental in the theory of disruption, which Christensen insisted on a lot, which is that the disruptive technology competes with what is called non-consumption. It will appeal to people who were not consumers of the current technology. For example, the first low-cost companies did not take customers from the traditional airlines, but from the bus lines. Before them, a traveler had the choice between slow but cheap (bus) and fast but expensive (plane). The low-cost airlines offered them fast and cheap, and they abandoned the bus immediately. Similarly, it was found that Uber customers were generally people who did not take cabs because these did not serve their neighborhood.

The lukewarm response of the incumbent

In summary, the incumbent being attacked will most likely respond with its own solution (which is what Google did), but it will not push it. On the one hand, most of its resources (financial and human) will remain dedicated to its legacy activity. On the other hand, its response will be to put disruptive technology at the service of its legacy activity. Thus, in 1996, Kodak released a digital camera… with a film! What was digital was the taking of the picture, which was then recorded on film. This reaction has a name: cramming. It consists in forcing the disruptive technology into the existing business model, rather than creating a new business model around the technology. Cramming has the effect of trimming everything that “goes beyond”, so to speak, i.e., all the disruptive aspects, in order to put the technology in continuity with the existing model. It thus ceases to be disruptive.

In order for a disruptive technology to become a real source of growth, it must therefore seek out non-consumers, i.e. it must not be forced to serve only the legacy model. In the case of Bard, Google would have to create a new autonomous subsidiary with its own entrepreneurial mandate, even if it means competing with the historical search business. This is unlikely, as the risk is too high given the considerable stakes involved (more than $200 billion in revenues).

The issue of substitutability

The incumbent player that is disrupted is not easily identifiable a priori. Spotify and dating sites have disrupted nightclubs, but who could have predicted it? It was not their intention anyway. It can happen that the disruption is directly substitutive. This was the case for Kodak: there was a complete substitution between film and digital, whereas there was only a partial substitution between, say, DVD and movie theaters. The question for Google is therefore whether conversational robots and the search engine are substitutable. If so, the danger is major. If not, there is at worst a loss of opportunity for Google in a new market. The answer is probably somewhere in between, as suggested by Microsoft’s approach of putting ChatGPT at the service of its Bing engine.

The innovator’s dilemma shows that while the incumbent generally responds to a disruption, it is constrained in its response for entirely rational reasons. It limits the potential of the disruptive technology by putting it at the service of its legacy activity, and by continuing to devote the majority of its resources to the latter. This leaves the field open to new entrants who do not have these constraints. However, Christensen points out that this is not a foregone conclusion, and that some companies have managed to escape the dilemma. It will be interesting to see if Google adds its name to this list.

➕Read my previous article Is Google Victim of the Innovator’s Dilemma with ChatGPT? On ChatGPT, read Assessing the Potential of ChatGPT: Lessons from the History of Innovation.

📬 If you enjoyed this blog post, why not subscribe?

🇫🇷French version of this article here.