Article Summary: Christensen's model of disruptive innovation tells us that when new market entrants offer products that are inferior to incumbents, at a lower price point, targeting non-consumers of existing products with solutions that are simpler, easier to use, more convenient and/or more accessible, that the upstarts will almost always win the competitive battle. We see this recurring pattern frequently, and it's easy to mistake these attributes as causing disruption, rather than as signals that disruption is happening. This two part series examines the root causes that enable disruption to occur.
Clayton Christensen, the father of disruption theory, has noted that "disruptive innovation" is a theory about who wins competitive battles when innovations are introduced to market. Specifically it is not a theory about how innovation works. (This single idea, by the way, should dispel all notions that any technology is inherently disruptive, that all innovation is disruptive, that if it isn't disruptive, it isn't innovation, or that any Silicon Valley startup that doesn't yet have a product ready for market could lay a claim to being disruptive.)
Christensen's model does not require any technology to be involved to create or enable disruption, although often technology does act as a catalyst, enabling new business models that do disrupt.
Where Christensen Gets it Wrong
Importantly, Christensen theorizes that the attributes that are part of the pattern of disruption are causal. So, you will often hear him say, or see written that disruption is caused when:
- a new market entrant introduces a product which is inferior to incumbents
- targeted at an un- or under-served market that incumbents find unattractive
- at a price that is lower than the existing market players
- is more convenient, easier to use, accessible, simpler
- using different channels or dis-intermediating channels altogether
- the new entrant is an industry outsider
- the product "competes against non-consumption" (this is actually a summary of a few of the above bullets)
But, does this pattern actually cause disruption? Clearly it cannot, as it is quite possible for an outsider (or an insider for that matter) to create an inferior product at a low price for an undesirable market that is more convenient or easier to use than existing products, yet have a complete market failure -- an innovation that no one wants or needs. This actually happens all the time. And, even if moderately successful, it's possible that the new product is useful only to a small low-end niche, and doesn't have an upgrade path that enables it to disrupt incumbents.
Examples of False Positive Patterns
Store Brands
An easy example is store brands, such as the tinned beans you purchase at the grocery store that have the retailer's brand rather than Van Camp's. These products may or may not be the same quality as the name brands, but we generally assume they are inferior (and minimally, they don't carry the same 'brand promise' of consistency, flavor, and support that the big name offers), they are lower priced (otherwise no one would buy them), and targeted at low-end consumers. There are thousands of store brands in every category, from food, to hardware, to sports gear, to clothing and so on, but are these products disruptive, and do they have any chance of ever disrupting the market?
The answer is of course, no. Store brands occupy a low-end niche, but never have and never will disrupt the big brand leaders. Niche products are not the same as disruptive products.
Dollar Shave Club: Also Not a Disruptor
Another example that is commonly touted as disruptive is Dollar Shave Club, yet is it? Even before being acquired by Unilever, DSC was not a disruptor, despite offering a lower-priced product via a different business model than Gillette (the market leader). Yes, they carved out a significant market share of 8%, and in combination with other low-priced offerings such as Harry's Shave Club and others, knocked Gillette from 71% down to 59% share. But seriously, can we say that the market leader who still controls nearly 2/3 of the market is disrupted?
DSC (and Harry's) have spent heavily on creative marketing programs and burned huge amounts of investor cash to buy those precious points of market share, but that can't continue, and at some point the investors demand a return. Does every new competitor who steals a few points of share constitute a disruptor if they have no chance to ever become the dominant market player? A couple of generations ago, Bic also carved out a low-end niche with disposable razors, but obviously never became dominant, and was also not a disruptor.
Remember, DSC, Harry's, Dorco and other low-priced brands together have only just bested the sleepy old Schick brand, which has done literally no marketing in a couple of generations. And, all the competitors to Gillette (including Schick, Bic, store brands, DSC, Harry's, Dorco, The Art of Shaving, and almost 50 more brands) combined still only total 40% of the market.
Moreover, consider that Gillette's home turf is being attacked, and that although they largely ignored DSC in the beginning, it is absolutely in their interest to defend their market, and the war chest to do so is immense. Whether by lowering price, offering a subscription service the same as the upstarts (which Gillette now does through Amazon, and through their website), increasing their marketing or creating alternative brands themselves, Gillette is highly incentivized to fight back, and the upstarts have neither a technology advantage nor a sustainable cost advantage.
In other words, DSC, despite being venture-funded and marketing their wares very aggressively, was not disruptive, and did not conform to the model of disruption (which says among other things, that there is an "asymmetry of motivation" between the incumbents and challengers when a product succeeds at disruption -- or, more plainly, the incumbent is motivated to flee a market rather than fight, while the challenger is motivated to attack). As noted, Gillette is highly motivated to defend their turf, and will absolutely do so, even if that means accepting lower profits for a while.
Again, remember that every innovator is not a disruptor, no matter how appealing or exciting their story is for a while. DSC carved out a niche, and they will probably gain a few more points, but their upper ceiling for market share is probably about 20% assuming they continue as strongly post acquisition, unless Gillette completely falls on their sword -- but then that would be attributable to incompetence and mismanagement, not market disruption. And, now that they've been swallowed by Unilever, they've lost their 'upstart' panache, and any chance of much further progress in gaining on Gillette is pretty much gone.
So, if the model is not causal (and therefore not predictive), what is it? We certainly agree that the model describes the definition of disruption, but the attributes are signals, or indicators, that disruption is happening or has happened. They are highly correlative, though not 100% (for the reasons described above), which is probably why Christensen and others have mistakenly viewed them as being predictive.
In my work, I have labeled these indicators as the "disruption fingerprint" for precisely this reason. Just like a fingerprint, they identify the suspect, but also like a fingerprint, these markers don't predict anything, except in a very trivial sense. What that also means is that if we want to create market disruption intentionally, we need to consider other factors -- the factors that create this pattern, not the pattern itself.
Ripples in a Pond
In my book, I use the analogy of ripples in a pond. The ripples are not the cause, but the effect -- the effect of a stone or some other object being tossed into the water.
“The ripple is evidence that a stone was thrown, an artifact of the event that created it. But, if you don’t know that a stone was tossed to create that pattern, you can’t reverse engineer it by studying the pattern and trying to recreate it by other means.”
Effect, not cause.
Another important issue with getting this wrong is that when relying on this imprecise shorthand, it also leads Christensen to misidentify some companies as being disruptive which aren't, and more importantly, to completely miss some of the biggest market disruptions that are obvious to almost everyone except the creator of the theory.
So, for example Christensen said the iPhone -- perhaps the biggest disruptive innovation in 50 years -- was not disruptive when it was originally released to market in 2007. (He later changed his mind.) Not because it doesn't actually conform to the theory, but because this overly simplistic shorthand doesn't consider that "disruptive" is a relative descriptor, not an absolute one, and critically, it depends on what market you are comparing to.
So, if you compared the iPhone to the cell phone market, then you would conclude that it was not disruptive, whereas if you considered it to be competitive to a mobile networked computer and compared it to notebook and desktop computers, you would have concluded the opposite (and we did). The example today is Uber, who Christensen and some of his other esteemed co-authors have labeled as "not disruptive", and again for the same reason: they don't understand what market Uber is defining -- importantly, while Uber is causing chaos in the taxi market, that is not the reason it is a "disruptive innovation". I will examine this more fully in a future article.
So, we conclude that Christensen's model defines what disruptive innovation is, but it is not causal and not predictive. But, that leaves us asking if not the attributes of Christensen's model, then what are the causes of disruption, and how can we use them to predict when disruption is likely and highly probable?
I will answer those questions in the next article.