Month: Dec 2015

In a recent “Critical Path” episode, Horace Dediu discussed the definition of disruption (00:46:40). Basically he outlines the boundaries of disruption theory and sets 3 criteria for Disruption.

1. Entrant product serves customers who are over served by current incumbent product, or serves non-consumers.
2. Asymmetry of motivation. The incumbent is not motivated to directly compete with the entrant.
3. There exists a “Technological Core” that enables the entrant to continually improve the product.

Now this is totally great. It is great that the criteria for a “Disruptive Innovation” is clearly laid out, and that we can now identify whether a certain upheaval in the market qualifies for the Christensen-ish definition of Disruption or not. I totally agree with the criteria, as should, I believe, anybody who has carefully read Christensen’s books.

The problem is, if you have three criteria that you can answer with yes/no, then you have 2 x 2 x 2 = 8 possible situations. Only one of these Christensen Disruption. What are the others?

Furthermore, it becomes important to understand what it means to be a “Disruptive Innovation”. Clearly, Christensen does not intend “Disruptive Innovation” to describe all situation where there has been a significant disturbance in the market. He only talks about one. Then the question is, what makes his single segment so important? More importantly, why are so many Venture Capitalists upset and why do they complain so much when a single academic declares that the companies that they are investing in do not fit his criteria of “Disruption”?

The key to this question is to understand what the power of “Disruptive Innovation” is. The companies that are “Disruptive” are the ones that can benefit from this power. The ones that do not qualify cannot.

In any market competition, the companies with the more resources generally win. The weaker companies typically lose. However, this isn’t much fun. Venture Capitalists looking for incredibly high returns, are instead betting on small and weak companies that will somehow accomplish something that much larger companies cannot. They are essentially betting on David beating Goliath.

Sometimes smaller companies will out-manoeuvre larger companies by being more nimble. Sometimes the willingness of smaller companies to experiment with crazy ideas will allow them to win. However, until Disruption Theory, there was no theoretical framework that could predict which smaller companies would have a high probability of success. And there still isn’t any other.

Disruption Theory is still the only well accepted business theory that has been demonstrated to be capable of identifying (probabilistically) the winners from the losers. It is the only theory that gives us a path that David could take to reliably beat Goliath. It is the only well known wave.

Companies that qualify as “Disruptive” in Christensen’s definitions are the ones that could benefit from “Disruptive” dynamics, and who can ride the waves to beat the most powerful incumbents. This means that they can make the most of the resources (capital) that investors pour into them.

On the other hand, companies that do not qualify must take a different path. Although each individual path has not been fully explored, weak entrants will typically not survive on these paths. There is not well identified wave to ride. There may be other waves, but we can’t reliably count on them.

I hope that this describes what not being “Disruptive” in the Christensen sense means. It means that if you are not “Disruptive”, you are not riding Christensen’s wave. If you are a battleship, then you are probably OK, but that means that there isn’t much fun for the investors who poured billions of dollars to build the battleship. Unless you ride the wave, you probably aren’t going to get insanely high returns.

Clayton Christensen has been working to limit the meaning of “Disruptive Innovation” to the definition the one that he uses in his books. By this, he aims to counter the absurd notion that “Disruptive Innovation” was supposed to explain every kind of market upheaval, hence its inability to do so suggests that the theory itself is invalid.

• “Fresh Insights From Clayton Christensen On Disruptive Innovation” Forbes
• “What Is Disruptive Innovation?” Harvard Business Review

Some people have suggested that since Christensen stressed that “Disruptive Innovation” is not everything, a new theory for explaining the other phenomena is required. For example, in the HBR article, Christensen said that Uber is not disruptive. Then what theory can explain the market upheaval that Uber is causing?

My suggestion is that we don’t actually need a new theory. All we need is a good understanding of what areas “Disruption Theory” covers, and a little bit of common sense.

Let me explain.

First, we have to start with a good understanding of exactly what “Disruption Theory” covers and does not cover, and to see what the theory omits. To do this, I will focus on what I think is the core tenet of the theory; how the incumbent will respond.

In “Disruption Theory”, Christensen covers two options that the incumbent may take. These are,

1. The incumbent responds directly to the entrant, and competes with it to drive it out, or to severely limit its penetration.
2. The incumbent decides that it is not worthwhile to compete with the entrant at the low-end, and focuses on the high-end. The word that Christensen often uses is “flee to the high-end”.

Both of these options assume that the incumbent is capable of responding to the entrant, and beating it if it so wishes. Although this may be more common, there is however little reason why this should always be the case. This is the omission. It is very possible that in some cases, the incumbent is simply incapable of responding.

It is also clear from a logic point of view, that adding this will fully cover the scope of possible options. The incumbent either chooses to respond, or not respond. In the case he chooses to respond, he either succeeds in launching a response, or fails. There are no other alternatives, no holes left open to consider.

In the following table, I have added this third possibility in pink. This is the possibility that the incumbent did not have sufficient resources to launch an effective response to the entrant.

On observation of this third option, it becomes clear why Christensen did not bother to spend time elaborating this option. It is just common sense. That is why I took the liberty of filling up the pink boxes with what I consider to be straightforward reasoning.

In this third option, the incumbent is unable to mount a response, simply because it is weaker. It might be a smaller company with fewer engineering responses. It might not have the marketing might of the entrant, or its brand appeal may be much weaker. It might lack the bundling opportunities that the entrant has. Or, in the case of the quickly changing tech landscape, it might lack the expertise that is required of the next phase of computing.

In fact, I would say that a better explanation of why Christensen failed to predict the success of iPhone is because he misunderstood how difficult it would be for Nokia to adequately respond. He underestimated the resources required to create an iPhone competitor, and the know-how that Apple (and Google) had accumulated as software companies. He overestimated Nokia, Palm and Blackberry’s ability to create a similar operating system, despite trying hard either in conjunction with Microsoft, or by bringing in Jon Rubinstein to head the WebOS effort. Christensen failed to correctly assess the huge advantage that Apple had in terms of resources relevant to smartphone R&D, and instead only evaluated the resource advantage Nokia had in distribution, etc. He failed to understand that the R&D resources required to compete in the post-iPhone world were not the same as what Nokia already had, and were instead abundantly available at Apple and Google.

If the entrant’s new product is well received by the market but the incumbent cannot create a product that is competitive, then the incumbent will lose. The entrant does not necessarily need to position its product at the low-end because it can win at the high-end. There is absolutely nothing surprising about this; it is simply common sense.

To summarise my table, I believe that “Disruption Theory” left out just one option, and that has caused confusion. By filling this in, I think I have fully covered the options regarding what an incumbent could do (luckily, the omitted option was common sense, and so filling it in was trivial). It is my belief that how the incumbent responds is critical, and very often the key turning point. This is why I do not think that it is necessary to further sub-divide each incumbent response category based on other facets like cost-structure, for example. I believe that the outcome will mostly be determined by the incumbent response, and that sub-divisions will not lead to substantially different outcomes, hence sub-division loses any practical utility. (This is to say that sub-divisions that only predict the same result as the parent category are not very useful. For example a sub-division inside the “cannot respond” category that predicts that the incumbent will fall is not really worthwhile, because the “cannot respond” category already predicts the same.) Furthermore, I doubt that a detailed analysis of the “cannot respond” category is necessary since it is already common sense for the vastly more powerful to prevail. I would however appreciate any discussion on any aspect of my table.

Finally, I would like to discuss what this means for Uber, the controversial example that Christensen touched.

My understanding of Uber is the following.

1. Uber is the entrant but is also the Goliath. As such, the Uber vs. taxi companies battle sits in the pink area of my table, the area where entrant Goliath predictably slays David.
2. Uber has accumulated billions of dollars in venture capital which it spends heavily on customer acquisition, driver acquisition, marketing and government lobbying. This in stark contrast to the typically small and not-very-profitable taxi companies, which have not accumulated vast amounts of cash, and cannot spend tons of money for the same purposes as Uber.
3. Uber is the world’s first global taxi company. It has benefits of scale in terms of its technology, branding and marketing.
4. Uber’s valuation is dependent on its not being seen simply as a taxi company, but also as a company that can disrupt many other industries. None of these efforts have yet been successful, but nonetheless, investors are seeing huge potential opportunities. However, taxi companies have never been seen in this light. They cannot raise capital with the same promises as Uber.
5. All this results in Uber being a Goliath in terms of the money it can spend, relative to local taxi companies. With vastly more resources at its disposal, it is able to invest in ways that its competitors simply cannot. Although there are no doubt other complicated reasons why taxi companies might have trouble launching an effective response, the enormous difference in cash cannot be ignored. Being Goliath, Uber will predictably slay the taxi Davids in the absence of regulation.
6. Importantly however, the fact that Uber is not “Disruptive” limits its ability to fulfil its investors’ dreams, even if it demolishes taxis worldwide. Uber is definitely a better mousetrap, but the economics does not ensure sustainably low prices nor sustainable profits. It does not reduce the burden of insurance, gasoline costs, car maintenance costs nor car depreciation. It does not eliminate the fact that the driver still has to make a living. These are the major costs of a taxi ride, and without significantly reducing any of them, there is little systemic reason why an Uber ride would remain cheaper than a regular taxi. As a result, the market will not significantly increase. Most people are unlikely to suddenly be enlightened to the possibility that ditching their own cars and using Ubers exclusively would be cheaper for them overall. This is what it means to not be a Christensen “Disruption”, but merely my pink box upheaval.
7. To be “Disruptive” to taxi industries, to bring taxi services to those who do not currently use taxis often, and to significantly expand the market, Uber has to either a) make taxis much easier to use for the people who currently have no access, or b) make taxis much cheaper. The first option is something that the green boro taxi program in August 2013 did in Brooklyn, with dramatic results. Applied to Uber, this would be like bringing taxis to rural areas where no taxi service is yet available. Option b) could be achieved by things like UberPool (which I don’t think is yet mainstream) or self-driving cars. UberPool is interesting and could fit the definition of “Disruptive Innovation” perfectly, if it worked. Of course, car pooling would compete more with busses than taxis, and we have to consider whether UberPool makes more sense. Self-driving cars are a bit too far out to sensibly evaluate.
8. There remains the possibility that Uber will be disruptive to other industries, for example, automobiles. This is probably the rationale for its astronomical valuation. I personally don’t think much of the prospects, but instead of confusing this post with another controversial discussion, I will leave this for another time.

UPDATE:

Cleared up the logic for Uber and added a few more details.

UPDATE 2:

We might want to consider cases where the incumbent changes their mind. An incumbent may initially dismiss the entrant as non-significant and decide to “flee-upmarket”. After realising that the entrant is a larger threat then they initially envisioned, they may choose to respond. Whether that response is too late or not will depend on how strong the entrant has grown. If the entrant has grown to the point that the entrant is more resourceful than the incumbent, then the entrant will win. If the entrant has not, then the incumbent will win. This is a situation that is well documented in Christensen’s work.

UPDATE 3:

With the pink category included, it is easy to see the large role that venture capital plays. Venture capital can instantly create Goliaths that dwarf the incumbents. Startups that have not accumulated any cash to date, can suddenly obtain resources many times larger than the incumbents’, if they can paint a picture that is sufficiently rosy to investors. Of course startups have to convince investors that the opportunity is much larger than the current market, much in the same way as Uber is pitching itself as more than a taxi company, but that is speculation-based, and startups do not necessarily have to provide hard figures.

In other words, the incumbents are at the mercy of the speculation that their industry could turn out to be but a small portion of a much larger enterprise. This does not always turn out to be true, as evidenced by how Google destroyed the RSS market (they obviously though that the RSS market would turn out to be something bigger that what it was), only to abandon it later.

UPDATE 4:

Just to clarify, I do not wish to debate the adequacy of Uber’s current valuation, nor do I want to denounce it as a VC bubble. I am totally uninterested in how Uber managed to convince investors of its valuation. I do not care if Capital follows Opportunity, or if Opportunity follows Capital. The only thing I am interested in is that the entrant Uber has vast resources while the incumbent taxi companies do not. This to me is an indisputable fact.

I do say that the taxi business does not validate its valuation, but I do not intend to question its valuation based on future, speculative opportunities. That risk is for the investors to take. They’re not spending my money.

Just a quick list of my thoughts on Disruption Theory;

What does Disruption Theory predict?

This is very important. This is the “reason for existing” of any theory.

1. Disruption theory is basically the prediction of how incumbents will react towards an entrant.
2. Disruption theory predicts how incumbents will react. The prediction is based on an understanding of how attractive it will be for the incumbent to flee upmarket. That is, to abandon the low-end market and focus on the high-end. Incumbents will typically only flee up market if their business as a whole is not threatened in the short-term.
3. If incumbents choose to react, and if the incumbents have huge resources relative to the entrant (which if generally the case), then incumbents are most likely to win.
4. Conversely, if incumbents do not choose to react, the entrants will gain a foothold at the low-end of the market.
5. If technological improvements and the profitability situation is favourable to the entrants, then they will progressively improve their product to the point that they will endanger, or even push out the incumbent.

So really, Disruption Theory does not directly predict success. It predicts how incumbents will respond, and it provides a general trajectory of how businesses will fare given their responses.

What is the predictive power of Disruption Theory?

The predictive power of Disruption Theory is similar to how the Newton’ Three Laws of Motion predict motion. If you are observing a curve ball thrown by a baseball pitcher, you need to consider Newton’s Three Laws, the Law of Gravity and aerodynamics. In particular, aerodynamics can be hard to predict and that is why we have air-tunnel experiment facilities (and no gravity testing experimental facilities). Disruption Theory is just one of these. Without good understanding of the other theories, the predictions of Disruption Theory will be inaccurate.

Furthermore, if the person applying the laws of physics to a curve ball does not know the speed of the ball, the angle to which it was thrown, the wind conditions, the rotation speed and more, then they will not be able to accurately determine its trajectory. This is despite the laws of physics being completely well known and defined.

Now the question is, is business more complicated or less complicated than a curve ball? If it is even comparatively as complicated as a curve ball, then it is reasonable to assume that even if Disruption Theory was 100% correct, we still wouldn’t be able to accurately predict the trajectory of a curve ball unless other laws were defined, and we knew detailed business conditions.

If you ask me, I’m convinced that business is more complicated than a curve ball.

Given the complication situation, the best we can hope from Disruption Theory is a probability percentage. This is like how we are given weather forecasts. Meteorologists don’t tell us exactly how the weather is going to be. They sometimes get it totally wrong. However, more often then not, they get it right and we are OK with that. Even despite the fact that the we fully understand the basic laws behind the weather, and have tons of data. It would be a monumental achievement if Disruption Theory ever reached the predictive power of a typical weather forecast. We can not reasonably expect more.

Faulting Disruption Theory because Christensen failed to accurately predict Apple’s fate, is like totally dismissing the basic laws behind the weather because a meteorologist made the wrong forecast once or twice. In short, it’s ridiculous.

Applying it to Uber

People in Silicon Valley seem to be very upset that Clayton Christensen said that their favourite company, Uber, was not “Disruptive” in his definition of the term.

Let me give my two-cents on the matter.

1. I agree that Uber is not “Disruptive” in Christensen’s definition. It is clear that Uber is directly challenging the taxi industry, and that the taxi industry is motivated to respond. In particular, the taxi-industry is in no way fleeing upmarket.
2. The consequence of not being “Disruptive” is simply that it will not follow the trajectory of what Christensen described. Incumbents will not flee upmarket and Uber will not drive them out through future improvements to its service.
3. Whether Uber will ultimately succeed is a totally different matter. They are free to pursue their own trajectory. However, what Christensen described is a way for a weak entrant to overcome a powerful incumbent. Uber will not be able to leverage the power that the “Disruptive Trajectory” provides. It will not be able to succeed as a weak entrant. It has to become powerful, more powerful than the incumbents to succeed.
4. Therefore, Uber will have to really on a brute-force approach. Fortunately, Venture Capitalists are happy to provide Uber with ample cash to spend on expansions and promotions. Uber as a result is able to use money-losing promotional tactics, something that the much smaller incumbents cannot. And clearly Uber seems too be compelled to use these tactics as they raise more and more money.
5. An important point to note is that taxi companies are small all over the world. They operate very regionally, most often with a territory that covers only a few cities. Often, taxis are even self-employed. Taxis are also not a very profitable business, and hence taxi companies have not accumulated enough capital to spend on high-tech, let alone self-driving cars.
6. Hence my understanding is that Uber is a Goliath entrant vs. the small incumbent Davids. Uber’s trajectory should be like how supermarkets obliterated mom & pop grocery stores. That is to say, you don’t need to know any kind of business theory, let alone “Disruption Theory” to see how it is going to end.
7. Hence the real question for Uber is not whether it can push out taxis. Without any regulation, the answer is clear even without “Disruption Theory”. The question is will Uber be a sustainable business? Will it raise prices after venture capital runs out and there is no competition left? If they are forced to employ their drivers as employees and if they have to also pay for their driver’s cars, which is quite possible long term, can they still maintain current prices? If Uber becomes a monopoly, will they be any better than the regulated monopolies before them for both the drivers and the customers? I have serious doubts on this, and unless Uber discloses the sustainability of its business, commits to future low prices and the welfare of its drivers, I think that strictly regulating Uber makes a lot of sense. The last thing that you want is for Uber to kill your local taxi industry, and replace it with one which is just as expensive (potentially more) and where all the profits are funnelled to a Silicon Valley company far away. This is why we have anti-trust laws, for example, and this is why we regulate industries (like the public transport, mail, health and food industries) that directly affect the welfare of our citizens.
8. Of course, there is discussion that Uber is more than just a taxi business. That it is a more general transportation/delivery service. This suggests that Uber does not compete only with taxis, but with busses, trains, subways, delivery trucks, Fedexes, Amazon drones, and more. Note the totally different economies regarding scale and costs, completely unattainable by people driving their private cars for Uber. Can Ubers compete with busses? Can Ubers compete with trains? Can Ubers compete with the Fedex delivery network? Can Ubers compete with Amazon drones? Can they ever be cheaper than these highly optimised systems, which use specialised vehicles but still aren’t too profitable to begin with? It’s easy to compare Ubers to taxis, but to the other businesses, not so much. It will be a much steeper challenge for Uber to disrupt these businesses, and even a phenomenally good mobile App is not likely to help much. Maybe it will work in regions with underdeveloped public transport infrastructure like California and some developing nations, but I don’t see it working like that in developed countries where auto lobbyists were not able to sabotage the development of public transport.

So in conclusion, Uber is definitely a threat to the taxi industry, but not because it is “Disruptive”. It is such a threat, probably just because it has the deepest pockets in the industry. It is a national supermarket network vs. a mom & pop store kind of battle.