— published = false

I’ll get straight to the point. I have very little patience for Clayton Christensen’s Theory of Disruptive Innovation. And this is not just because the term “disruption”, which Christensen used to characterize a well-defined, academic theory (one of the few business theories I might add that has a body of published evidence to support it), has been used in contexts which wildly outstrip the boundaries of the original definition. And it’s not necessarily because I think any single premise in Christensen’s theory is fundamentally incorrect. In fact there are several anecdotes about the process of technological change that Christensen discusses or touches on in his books and articles which show up persistently accross numerous examples of technology-driven businesses that I can find. Rather, what I will posit in this article is that Disruption Theory is not a complete theory of innovation. That is, it does not fully encapsulate the many myriad ways in which new product categories, often (but not always) commercialized by new ventures (read: startups) can create large, important businesses, sometimes at the expense of large, well-funded, and well-managed competitors. On the other hand, I will argue that it describes one “vector” or process of innovation among many. And in arguing this, I will also attempt to give you the outlines of my own framework for parsing the dynamics of technology-driven industrial competition that I’ve cobbled together with (very few) of my own insights, and (many) other individuals’ hard work and frameworks.

Disruption Defined

Before I tell you why I don’t like disruption theory, I should probably tell you what disruption theory actually is.

Essentially, the narrative goes like this:

1. There is a firm (“the incumbent”) which makes a product category that is sold to a group of existing customers to solve a problem or meet a need/want. In doing so, this product category meets that need by competing along a set of performance dimensions, e.g. storage capacity (in MB), speed, fuel economy, etc.. These customers are aware of the problem they need solved, compare amongst the competing products within the category and between it and its substitutes, and pick the product for which the performace along a set of dimensions best meets their needs.
2. These companies, being generally well-managed firms, understand these customers and their needs/wants, very well. And so when technological innovations are created that promise to improve the product categories’ performance along the performance dimensions that are desired by the firms’ existing customer base, the incumbent jumps on the opportunity to better serve its customers and thus is generally very good at spending the resources necessary to incorporate these innovations into the next generation of the product category, and speeding the product into the market. Innovations of this type are called sustaining innovations.
3. Product categories built around disruptive innovations on the other hand, tend to be inferior along those same performance dimensions that are valued and needed by an incumbent’s existing customer base, and thus are rejected by the customers the incumbents have served in the past. But, they have unique advantages along one or more other peformance dimensions. In The Innovator’s Solution, Christensen states that Disruptive Technologies, “are not as good as currently available products. But disruptive technologies offer other benefits — typically, they are simpler, more convenient, and less expensive products that appeal to new or less-demanding customers”. In the setting of “New Market Disruption”, these features are said to: A. Enable “a whole new population of people to begin owning and using the product”, or B. Allow the product so operate in new applications (i.e. solve different problems than the incumbent product category). That is, although these products would be dead-on-arrival in the incumbent’s main markets - those well-served by existing product categories - they find new markets to serve, either in the form of people who would like to use incumbent product categories to solve their problems, but can’t, or by using their unique performance attribute(s) to solve new problems for which there were previously poor solutions. For individuals who were previously unable to access the older product category, the “good enough” nature of the new product category is better than nothing, and in new problem-settings, unique features of the new product category make the product worth it, while the performance dimensions in which the new product category underpeforms are, again, “good enough” to suffice for the new application. In this sense, they “compete against nonconsumption”. The examples that Christensen uses to describe new market disruption of this type are the Personal Computer, Sony’s first transistorized radio, and the Canon Desktop photocopier.*
4. Then, as the disruptive product category improves along the performance dimensions (if possible to do so; sometimes, Christensen points out that it is not) that are actually valued by the customers who are served by incumbent product categories, “they ultimately become good enough to pull customers out of the original value network into the new one, starting with the least-demanding tier.” The new product category “pulls” customers from the old product category to the new one, leaving the old one to collapse and die.

There are several characteristics of this process that Christensen heavily emphasizes.

1. In the setting of new market disruption, the fact that the new, disruptive product category first serves those individuals who are not considered part of the incumbent’s main market is extremely relevant to one reason why new entrants can succeed at the expense of otherwise advantaged incumbents. The reason? Firstly, these “new” markets are unknown - since the customers are by definition not those the incumbent mainly serves and thus not the ones the firm has built its internal processes to serve well. That unknown nature entails a degree of unquantifiable risk in developing for these customers - a tenet that goes against well-run company’s strategies for filtering out investments that are not likely to meet customer demand and thus have a positive return. Normal questions asked before product development: Do the customers want the product? How big will the market be? Will tbe investment be profitable?” are very hard to answer with any degree of certainty. A famous Christensen saying is, “markets that don’t exist can’t be analyzed”. Because of this uncertainty and risk, incumbents just don’t bother bringing to market disruptive product categories, leaving the field clear for new entrants. Secondly, even when analyzed, these markets tend appear small and unprofitable relative to the incumbent’s main markets and thus incumbents are not motivated to make the resource commitments necessary to create a product for these new customers. Note this focus on “resource allocation”. Christensen’s ultimate argument is that the qualities of disruptive product categories and the customers they serve - simple, cheap, low-performance products targeting new, unserved, unprofitable customer - mess with incumbent mental models about how to compete by deliberately violating the tenets of product development that work for an incumbent’s main set of customers. He provides then, an internally consistent (if not entirely correct in my opinion) mechanism to explain why a new entrant can succeed in the face of the advantages of an incumbent.
2. Christensen almost always notes that disruptive technologies proceed by targeting previously unprofitable (i.e. low-margin) customers by using products that are cheaper, simpler, and “good enough” relative to incumbent product categories. That is, Disruption proceeds always from the “bottom up”, meeting customers with low-needs first, then expanding in capabilities to pull in customers who are ‘higher-margin’ with higher peformance demands. This fact , referring back to point #1 above, is a large part of the reason why these “new markets” seem unprofitable to large incumbents. And in fact this “vector” of attack for a disruption is critical to the reason why new entrants can succeed at the expensce of incumbents. If, for example, a new entrant introduced a disruptive product category that was immediately, off-the-bat better along an existing set of performance dimensions, then Christensen predicts that an incumbent would immediately identify the product category as a threat to its offerings and devote the resources necessary to bring it to market. Thus for Christensen’s theory to work, disruption must be a “bottom-up” process.
3. In The Innovator’s Solution Christensen changes his focus from disruptive innovations to disruptive business models, arguing that “few technologies are intrinsically disruptive or sustaining in character; rather, it is the business model that the technology enables that creates the disruptive impact.” But the properties of this business model are consistent with the description I’ve stated above. Christensen himself states that new market disruptive innovations must have a business model that can, “make money at lower price per unit sold, and at unit production volumes that initially will be small. Gross margin dollars per unit sold will be significantly lower.”

What I Think Disruption Misses

So that’s a basic introduction to disruption. If the theory sounds well put-together and internally consistent with itself, that’s because it is. And in fact, before I begin my critique of the theory, I want to note two elements of technological history it gives robust explanations for and as such, any modification to the theory (including my own) will have to come up with alternative explanations for.

Firstly, as mentioned above, the theory provides a convincing, structural explanation for why startups are often at an advantage when it comes to commercializing certain new product categories. This explanation is actually one of the reasons the theory is so popular, particularly amongst VCs, founders, and other individuals involved in the startup ecosystem. Marc Andreessen of a16z once mentioned on a podcast, “Tons of tons of books will tell you company X screwed up beacuse they were incompetenet and dumb. To which the lesson apparently from that is ‘don’t be incompetenet and dumb’, which is, I guess a useful peice of advice. The thing with Clay’s book that was so strikingwas the observation that when big companies get disrupted, it’s not because they’re incompetenet and dumb. In fact quite, the opposite. It’s beacuse they’re very competent and smart…And for me that was a big mental left turn.” (https://a16z.com/2017/09/01/disruption-jtbd-modularity-christensen/)

And indeed, this ability to give a reason for startup success that is fundamentally not dependent on big-companies being incompetent is vitally important to any theory of innovation because as Christensen, Marc, and many others correctly point out, big companies are not incompetent most of the time. Yes, they have some aspects of bureacracy, quarterly focus, poor-planning, and a “rest-on-my-laurels” attitude. But mostly, they’re exceedingly well run, know their customers’ needs and wants in exquisite detail, and spend billions on research and development specifically for the purpose creating new innovations that help their customers. And thus, a theory that explains why

Secondly, disruption hones in on one other aspect of technological history that’s also readily apparent: many important new innovations and their associated product categories - the ones that end up having a disrporportionate impact on history, society, and business - often start out looking like toys and are thus derided for being unimportant. They’re often crude first-attempts, not-at-all refined in comparison to the products that customers might otherwise choose. Alfred Chandler, another eminent business historian, talks about the idea of learning bases, which he defines as a firm’s “organizational capabilities involved in improving existing products and processes and for developing new ones in response to changes in technical knowledge and markets”. These new product categories have no such learning bases to draw on, whereas the incumbent firms, making incumbent products, do. Correspondingly these products are refined and effective in comparison to the new product category. Cars, cell-phones, smart-phones, airplanes, the original telephone, original radios, etc., all seem to suffer from this “innovator’s curse”, whereby no one seems to take them very seriously when they first arrive on the scene. Disruption theory argues that this feeling we have - that innovations seem “hacked-togther”, unrefined, feature-light, and almost “toy-like” when they first arrive on the scence - is not a figment of our imagination, but very important reasons for why startups can commercialize them at the expense of large incumbents.

Both these two aspects of Christensen’s theory - explaining structurally why startups have a fighting chance in the face of incumbents & explaining why new innovations often seem like toys - will be crucial to my own extension of the theory.

Disruption and History

But if I like these kernels of truth that Christensen unearths with his theory, why, then, do I not like the entire theory itself? Well, I will aim to prove that disruption does not fare well when analyzed against the glaring light of historical record. Christensen himself has stated, and I’m sure many others will agree, that the measure of good theory is not whether it’s internally consistent or not (Disruption is!), but whether it accurately describes reality. And when I look at the record of new companies commercializing product categories based on innovation, while I do see many which followed Christensen’s process accurately, I see many more- the most important that I’ve read about - which do not meet Christensen’s rigorous definition of disruption. Now, Christensen has made misteps in the very-recent past related to disruption theory, famously when it came to the iPod and the iPhone, as well as with Uber; each of these product categories clearly is not “disruptive”, but has been wildly successful. See this article by Ben Thompson for more. But I want to focus on other examples that may not be as well known and do not have the unique characteristics of modern “marketplace” companies, hoping to show that disruption cannot accurately describe the industrial dynamics of the last century’s most important products. To this end, I begin with one of America’s most famous corproations: AT&T and Bell’s eponymous telephone.

Example 1. AT&T vs. Western Union

In one of the first pages of “The Innovator’s Solution”, Christensen states that, “AT&T’s wireline long distance business, which disrupted Western Union, is being disrupted by wireless long distance”. This statement - one seemingly innocouous sentence in a 300 page book - caught my attention because I previously conducted quite a bit of intensive reading into the early financial history and competitive strategy of AT&T. As one of America’s first durable business franchises funded by private capital, AT&T in a way prefaces the history of modern venture capital and I wanted to understand the lessons its rise could have for building durable, dominant companies, particularly in so called “hard-tech” areas. But as a side-benefit, I knew immediately that the telephone - an example of a underfunded startup with a new product category displacing an incumbent - was definitely not disruptive.

The telephone was concieved and marketed as a 1-1 replacement for Western Union’s system of telegraphs.

Example 2. Electricity

Example 3. Transistors and Fairchild; Microprocessors and Intel

Example 4. Medicine and mABs

Example 5. Synthetic Dyes

““The Ford Model T, for example, created the first massive wave of disruptive growth in automobiles.”

““AT&T’s wireline long distance business, which disrupted Western Union, is being disrupted by wireless long distance.”

• Note that to support this definition, Christensen argues that 2 ways to identify “new market disruption” are to ask: “Is there a large population of people who historically have not had the money, equipment, or skill to do this thing for themselves, and as a result have gone without it altogether or have needed to pay someone with more expertise to do it for them?” and “To use the product or service, do customers need to go to an inconvenient, centralized location?”