Driving Disruption
“Upward, not Northward”
― Edwin A. Abbott, Flatland: A Romance of Many Dimensions
The current fitful, disruptive AI era, discussed in my last post, is a case study in Clayton Christensen’s business theory of innovation that’s unfolding as we watch. Christensen’s Innovator’s Dilemma helps explain why incumbent market leaders can’t seem to jump on the new disruptive innovation but, instead, continue with sustaining innovations that ultimately lose to new market insurgents.
A great example is Kodak missing the gigantic digital photography market, even though they invented the technology way back in 1975. Apple, Google, and Facebook clearly won the digital photography market by making it a killer feature in their products, fueling nearly $2 trillion in market value between them.
So what are the mechanics of all this? Stay with me, it might get a little nerdy.
Typically, a new product (green curve, Figure 1) enters a market with poorer performance than the incumbent product (purple curve, Figure 1). But, the insurgent gets better over time and fuels its growth by finding a new market that the incumbent is underserving, or that’s enabled by some new technology.
For example, in the 1960’s, the thrifty Corona and Corolla made car ownership possible for those who couldn’t afford a car, and generated huge market growth for Toyota in the process. [My first car was a slightly used Corona, paid solely by my wages as a Miami busboy. It was silver, and it was awesome.] By 1975, Toyota had surpassed Volkswagen to become the #1 import brand in the United States.
Once the incumbent realizes the insurgent is progressing, it reacts by continuing its habit of incremental, sustaining innovations (the dotted purple line, Figure 1). This may be necessary to maintain its current market position for awhile, but is insufficient to catch the insurgent. In Toyota’s case, the successful Corona and Corolla insurgency laid the foundation for the company’s growth to become the world’s largest automaker.
The problem, which is difficult for the incumbent to see, is that the insurgent is innovating in dimensions that may be completely orthogonal (i.e., unrelated) to the incumbent (Figure 2). These new dimensions could be new technologies, business models, or regulatory regimes. Insurgents are also culturally adept at operationalizing these new dimensional advantages because they’re smaller, faster, and have a lot less to lose.
If the incumbents were Abbott’s protagonists in Flatland, they’d go inwards to catch the green S-curve in Figure 2, not upwards on the same purple S-curve. They may be more comfortable on the purple S-curve, but they can’t win by staying on it.
One way to make the jump to a new S-curve is by being humble enough to align with the insurgents through partnerships, investments, and acquisitions (Figure 2). These strategies, while far from easy, are critical to making the jump. High tech companies, like Cisco and Intel, have been running parallel experiments through their startup investments for decades, acquiring the winners, and jumping to the next disruptive S-curve. Now automotive leaders, like Toyota, are recognizing that they may need to do the same to compete against new high tech insurgents.
A good example of tapping into disruptive innovation can be found in the original content streaming market. About 10 years ago, it was becoming clear that video and music content would be delivered by streaming media, rather than static storage like DVD or MP3. Content distribution services, like Netflix, realized that it would be much easier for original content producers, like HBO and Disney, to stream their own content and fatally disrupt Netflix’s business. When asked what justified Netflix spending billions on original content production, they famously replied that, “the goal is to become HBO faster than HBO can become us.”
This disruption of original content delivery was not lost on Amazon, which realized quite early that, in order to cement its competitive position, it needed to jump on this new S-curve through more than a dozen partnerships, investments, and acquisitions. Amazon’s larger acquisitions included the 2014 acquisition of Twitch (a live video streaming platform for $1.1 billion, ~0.7% of their market value at the time) and the 2011 acquisition of LOVEFiLM (a UK film download service for $315 million, ~0.4% of their market value at the time). Clearly, Amazon needed to become Netflix and HBO faster than either could become Amazon.
Integrating acquisitions are not easy, often messy, but they can work. The biggest failure modes I’ve seen are misaligned incentives and incompatible cultures. Incentives are easier. Cultures are harder. Sure, the acquired team must have financial upside. But what’s harder is insuring that the acquired team has the agency, shared mission, and social connection to the acquirer. That’s a lot of risk for the acquirer, especially at high acquisition prices.
One way to mitigate acquisition-integration risk is through startup investments that help the incumbent learn how these disruptive teams are incentivized and operate. For example, the automotive industry is not data native like many high technology companies — e.g., Google, Netflix, Facebook, and Amazon. Traditionally, cars have thrown off data exhaust for, say, vehicle testing. Only recently has data fueled the creation of automotive systems, like automated driver assistance and fully autonomous vehicles. In Figure 2, data nativity is a “known unknown” that’s critical to the automotive industry jumping to the next innovation S-curve. Startup insurgents, not automotive incumbents, are driving this data disruption.
High technology companies have learned to disrupt themselves into the next discontinuous S-curve of innovation. The automotive industry is now learning to do the same. Toyota’s president recently said, “The automotive industry is currently facing a big change, what we thought was the future may just happen tomorrow. We need to be attacking and defending at the same time.”
The era of disrupting while sustaining has begun. Can automotive become high tech faster than high tech can become automotive?
