A Blog by Jonathan Low

 

Jan 29, 2016

The Stack Fallacy: Why Big Companies Keep Getting Disrupted

It's not the business model, or the coding, or the compensation scheme, or the hardware: it's the user. JL 

Christopher Mims comments in the Wall Street Journal on Anshu Sharma's article in Tech Crunch :

Many businesses that should have the resources required to build the next big thing often fail to do so. The “stack” is the layer cake of technology that ultimately delivers a product or service to the user. The reason that companies fail when they try to move up the stack is they don’t have empathy for what customers of the product one level above theirs in the stack actually want. In terms of who wins in a given market the question is who understands the user better?
If you work in tech and your company isn’t headquartered under a rock, you’ve heard of Clayton Christensen’s book, “The Innovator’s Dilemma.” You may have even thought a great deal about how his concept of “disruption” applies to your own firm, as either the disrupter or, more likely, the disruptee.
Whatever its explanatory power, the one thing “The Innovator’s Dilemma” can’t explain is why, almost 20 years after the book was published, so many companies continue to fall prey to more nimble competitors.
Anshu​ Sharma, ​ a venture capitalist at Storm Ventures, thinks he knows why so many companies that should have all the resources and brainpower required to build the next big thing so often fail to. He calls his thesis the “stack fallacy,” and though he sketched its outline in a recent essay, I found it so compelling that I thought it worth a more thorough exploration of the implications of his theory. What follows is the result of that conversation.
“Stack fallacy is the mistaken belief that it is trivial to build the layer above yours,” Mr. Sharma wrote. And as someone who worked at both Oracle ORCL 1.20 % and Salesforce, CRM -2.58 % his exhibit A is these two companies. To Oracle, which is primarily a database company, Salesforce is just a “hosted database app,” he wrote. and yet despite spending millions on it, Oracle has been unable to beat Salesforce in Salesforce’s core competency, notably customer-relations management software.
It helps to understand that in tech, the “stack” is the layer cake of technology, one level of abstraction sitting atop the next, that ultimately delivers a product or service to the user. On the Internet, for example, there is a stack of technologies stretching from the server through the operating system running on it through a cloud abstraction layer and then the apps running atop that, until you reach the user. Even the electricity grid required to power the data center in which the server lives could be considered part of the technology “stack” of, say, your favorite email service.Samsung,
originally a maker of components for Apple, tried to move up the stack by creating its own cellphones. At first it succeeded, but lately it has foundered on its inability to differentiate its products by creating its own software and is now increasingly overrun by competitors taking advantage of the commodification of Android smartphones. IBM IBM 1.04 % moved up the stack from making things that compute to selling the services that computation enables, and the result is that its revenue has shrunk for the past 15 quarters. Google tried to move up the stack from search to social networking and the result was Google+, about which the less I say, the better.
The reason that companies fail when they try to move up the stack is simple, argues Mr. Sharma: They don’t have firsthand empathy for what customers of the product one level above theirs in the stack actually want. Database engineers at Oracle don’t know what supply-chain managers at Fortune 500 companies want out of an enterprise resource-planning system like SAP, SAP -1.51 % but that hasn’t stopped Oracle from trying to compete in that space.
To really understand the stack fallacy, it helps to recognize that companies move “down” the stack all the time, and it often strengthens their position. It is the same thing as vertical integration. For example, engineers of Apple’s iPhone know exactly what they want in a mobile chip, so Apple’s move to make its own chips has yielded enormous dividends in terms of how the iPhone performs. In the same way, Google’s move down its own stack—creating its own servers, designing its own data centers, etc.—allowed it to become dominant in search. Similarly, Tesla’s move to build its own batteries could—as long as it allows Tesla to differentiate its products in terms of price and/or performance—be a deciding factor in whether or not it succeeds.
Of course, the real test of a sweeping business hypothesis is whether or not it has predictive power. So here’s a prediction based on the stack fallacy: We’re more likely to see Uber succeed at making cars than to see General Motors GM -0.65 % succeed at creating a ride-sharing service like Uber. Both companies appear eager to invade each other’s territory. But, assuming that ride sharing becomes the dominant model for transportation, Uber has the advantage of knowing exactly what it needs in a vehicle for such a service.
It is also worth noting that the stack fallacy is just that: a fallacy and not a law of nature. There are ways around it. The key is figuring out how to have true, firsthand empathy for the needs of the customer for whatever product you’re trying to build next. Apple, for one, spends a great deal of time and effort testing the products it rolls out, to the point that it is rarely first to market in a category. That is one reason its attempt to move up the stack from cellphone components, including chips, software, batteries, and advanced lightweight materials, to the next object to use them—electric cars—might succeed.
In terms of who wins in a given market, says Mr. Sharma, the fundamental question is and has always been, “Who understands the user better?”

1 comments:

DavidLJ said...

Note that testing hypotheses by seeing whether their predictions pan out is resting your life on a fallacy. You're trusting yourself to an affirmation of the consequent.

This is the failure to fail. It may, but not must, be thought of as allowing a measure of plausibility.

Faiure to fail may allow success. Ensure it?

No.

-dlj.

Post a Comment