Game Changers II

Disruptive Innovation

Short Recap

In the last post I talked about disruption that is “invisible”, in that it happens in a dimension that seems unimportant or unlikely to yield real game changers. Take something boring, like manufacturing process, add something small that starts to make slow and incremental improvements in it, wait a long time, and without ever being able to say where the change happened, the operating environment (“OE”) has undergone a dramatic change that has left previous incumbents in tattered ruins and new upstarts in glorious palaces.

The reason even very smart people with incredible business or strategic leadership skills, like Andrew Grove of Intel, find it difficult to see this kind of “invisible” disruption, even in hindsight, is because it all happened in a dimension that everyone was already occupying and just not paying a lot of attention to. 3dfx employees always assumed (so the apocryphal tale goes) that Nvidia had some kind of secret sauce, and discovered to their dismay when Nvidia bought them out that it was just efficient operations. 3dfx had operations, how is it that they could be so completely wiped out in a couple of years from super-dominance to being bought out? Well, this business equivalent of being taken by complete surprise was because they were focusing on the wrong strategic dimension. In their case they thought the game was being played in the the “how clever and technologically advanced is the chip” dimension, as opposed to “what can we do to match the development cycle of the machines these chips slot into” dimension.

The Techy Kind of Disruption

Today’s topic is the other kind of disruption, the kind the Nvidia is better known for now and the kind that (in a different guise) is much-studied in business schools. In the language of the generalised strategic analysis framework developed in these posts, this is the kind of disruption that is caused by a change in the conceptual part of the OE (“wait, could bishops go straight if we give them tanks?”) rather than a change in the conditional part of the OE (“is it snowing during today’s match?”). The typical example in OEs that are both unstable and dynamic (like business or warfare) is disruption through technological change.

The first thing to note is; this kind of disruption is only possible in environments that are not inherently stable. If the game or other strategically relevant situation (“SRS”) is only possible if the rules are rigid, no amount of technological development will get you to a rule-change in your favour.

This looks like just definitional language games; stable OE means rules don’t change. But people have a hard time grasping this when it occurs in an actual SRS. For example, I have had to make this point in the past with litigation strategy clients. There just are certain procedural rules that you must follow, otherwise you are not litigating and the court will not hear your case. Those rules may sometimes seem daft, but they exist precisely so that you cannot disrupt the basic dimensions of the SRS. If you want that kind of dispute, you can take the dispute into a different domain where the OE is less stable, e.g., you can try to beat your competitor in business rather than in the courtroom.

Now, how does this kind of disruption take place? Let’s take some famous examples: Walmart disrupted K-Mart; Amazon disrupted, well, pretty much everyone,

We humans love coming up with rules for everything. Pick a profession. Go ask someone in that profession how to become a member of it. They will list a bunch of informal rules, not all of which are useless but all of which are weird. You want to be a barrister practicing commercial law or tax? Well, you need to study the six foundational subjects (in law school or in a conversion course after your actual degree), get a first class degree, do a few work experience programmes, get into Oxford for their BCL programme and do well, join an Inn of Court and have 12 dinners there (I’m not kidding!), apply for a scholarship and ace the interview so your Inn pays for your Bar Course (whether you have money or not), pass your bar exams, obtain a “pupillage” (first year of paid work when you are being guided by more experienced barristers), obtain a practicing certificate and get voted in to join your chambers as a “tenant”. Some of these are more formal requirements, some are completely informal but practically necessary, and some (like the dinners) started out as informal but became formal rules over time.

The same is true of all kinds of human endeavour, not just applying for jobs. Entire industries have formal requirements (like licensing or regulatory requirements) and informal “this is the way it is done” -kinds of rules. Over time these rules and other behaviour-guiding knowledge become conceptualised and made into models and frameworks, that are easier to follow in corporate and other aggregate behavioural contexts (like professions, guilds, authority hierarchies or supply networks) but hard to unpack into their original constituents. Before long, nobody knows or thinks about why something is done the way it is, because nobody needs to (and therefore it would be inefficient to do so). Pressure to do things differently mounts in the OE; the OE becomes charged, until someone discharges it by breaking the model.

In unstable OEs where the existence of boundary rules does not define the OE, the “rules” of the SRS (conflict, game, or whatever) are more about what you literally can or cannot do in order to reach a certain goal. Finding ways to reach the goal without following the rule is the second kind of disruptive change. If someone realises that you can in fact shoot arrows while riding a horse, he may just be Attila the Hun. If someone realises you can in fact be a discount retailer without having storefronts, he may be Jeff Bezos. In business school literature, this is called “disruptive innovation,” often contrasted with “sustaining innovation” (or innovation that can be used to do things in the same way as before, but better, more profitably or more efficiently).

What Does Disruptive Innovation Disrupt?

Disruptive innovation is easier to grasp with an example. Let’s imagine a frictionless world where the stuff I said about corporate operations and their structural issues in the last post don’t exist.

So an iPhone comes along. So what, says Nokia, the incumbent mobile phone giant; we have touch screen tech, might take us a bit to hit the design window and get the supply chain issues sorted but they should roll out of the factories in a year or two. This, in the business school lingo, is known as a neutralising innovation. Microsoft did this with Word, Excel, Internet Explorer (in its memorable duel with Netscape) and many others. The product doesn’t have to be best in class or even all that good to begin with, it just needs to exist so your customers don’t flock to your competitor while you work on improvements.

So why didn’t Nokia do this? Well, the answer is they did. But the phone design isn’t really what the iPhone disrupted. Although the technology is cool and the design sleek, the real innovation is a business model change. The phone is no longer just a device for calling and texting, but a platform for all kinds of apps made and monetised by a developer network willing to pay Apple rent for being able to get on the platform. The user gets access to richer information that doesn’t need to pay the richness-reach tradeoff in the same way as before and more functionality that grows by the day. In return, the user gives up data that can then be harvested and used to build new network effects to attract more users and eventually even to build new products.

(SIDE NOTE: the Richness-Reach tradeoff is roughly this: you can tell your entire life story to your friend at a pub, or you can say hello to millions of people in a short ad in the New York Times, but you can’t tell your life story in the NYT. You can give rich information to few, or reach a lot of people with less-rich information, but normally not both).

The business model disruption effected by the iPhone means that if you try to compete with Apple by just selling phones, you lose. Apple makes money selling the phones, but also takes a cut off the App Store revenue, sells Apple TV and Apple Music subscriptions, and can refine the product offering to reach new customer segments with new iPads, Apple Watches, AirPods and so on. All the while, they gather data that makes these product offerings more and more directed and can gradually hike up the price of an iPhone because its functionality grows without Apple having to pay for it (in fact, Apple gets paid when new apps are sold on its platform).

Business Model Disruption—How is This Still a Thing?

Now, add friction back to the world. Everyone who has seen the insides of a tech company knows that over time they tend to organise around their product offerings. Let’s say the product is a computer; this team builds the central processing unit, that team designs the chassis, a third team works on the screen, another on the operating system and so on. Eventually these parts get put together and made to talk to one another by a protocol team, even though it could have been made much more elegant if it had been designed as a single machine. Product design decisions that do not respect these boundaries might get the amber or the green light, but since there is no cross-functional team that would have decision-making authority over the various product teams and who would actually design and build the new product category (or borrow development resources from other teams), the idea slowly withers away and dies on the floor of the operations room.

Take a look at this bit from The Soul of a New Machine, where Tom West of Data General is racing to design a disruptive mini-computer against the incumbent Digital Equipment Corporation (or DEC) who have just come out with their entry, VAX, into the same product category:

“Looking into VAX, West had imagined he saw a diagram of DEC’s corporate organization. He felt that VAX was too complicated. He did not like, for instance, the system by which various parts of the machine communicated with each other; for his taste, there was too much protocol involved. He decided that VAX embodied the flaws in DEC’s corporate organization. The machine expressed that phenomenally successful company’s cautious, bureaucratic style.” (Emphases added.)

What is going on here is that companies (tech and otherwise) have a product development model, that slots in with their operational model and, ultimately, their business model. Often, perhaps most often, these models are not consciously designed or not in keeping with the expressly stated “official” model. I have heard this first hand from businesses I have consulted with. I once suggested that a part of the work force who were literally idle during certain parts of the development cycle could be rented out to external clients without any leakage of important proprietary information. The answer? “We have agreed that this is not a scalable business model”. This still confuses me. The company business model dictated that developers and other staff stand idle because making money from small software projects on the side didn’t scale?

I could speak at length about corporate and other organisational structures and how they operate when something outside the accepted model hits the scene. And maybe I will, but not in this post. What is important for now is to realise that participants in SRSs (businesses, opponents, partners, players etc) can often see the technology or other innovation with potential to disrupt strategic models but still won’t, or can’t, do anything about it.

This is because of marginal costs and payouts. Eating the development cost on the new disruptive product category or other game-changing innovation is simply not how we make money in this company, is the answer you will hear from a cadre of middle managers, even, or especially, if they are any good. Their job, after all, is to supervise the flow of company resources and ensure that it goes into profit-making enterprises that deliver. They will not back projects that might produce a game-changing product category five years from now; they may not even work at the company in five years, and before that they have salary reviews to worry about.

Same goes for shareholders. Their interaction with the company is informationally poor; they mostly have no idea what is going on inside the company, and mostly only pay attention to the share price. The share price is a single dot that is meant to represent a vastly rich information landscape ranging from future plans, production efficiency, corporate culture, how well Janet and Chris from accounts get on, new product marketing approach and so on. It is tragically easy for the process of abstraction from all this data to a single number to go horribly wrong and completely misrepresent the importance of one aspect or another. However, one thing is clear; investing in the development of a disruptive innovation will do nothing but bad things to the bottom line, and therefore the share price, no matter how compelling a case the CEO can make for this development in more information-rich environments (like over a couple of one-on-one pints in a quiet pub).

Don’t believe me? Ask Nvidia about their development of CUDA and building up the developer network around it. This was the defining stroke of genius by Nvidia that led to its market leadership in AI and other general purpose computing using GPUs, is responsible for their most upmarket products and most of their profits. But at the time when CUDA and the network of academics, developers and other relevant participants was nothing but a costs sink, Nvidia took enormous amounts of shit for it and had to fight off activist hedge funds and shareholder revolts.

Furthermore, established and well-run companies cannot disrupt themselves without destroying the thing that makes them the incumbent. They have supply networks, staff that is far along the experience curve (i.e. can make stuff quickly and efficiently), customer relationships, product lineups, market share; the stuff that my startup clients dream about having one day. This is how the business model is often entirely unintended, or certainly unintentionally rigid. There is tremendous value in the assets of an incumbent, and only a fool would voluntarily destroy it when others are working so hard to gain these assets.

For these and other reasons investing in disruptive innovation is often simply not a good bet for established companies; they make plenty of money already, their profit margins are good, why try this new thing that will probably just eat a lot of costs and not produce anything tangible? Leave that stuff to the academics, and hike up the price of the cash cow product while we’re at it. By the time the technology or other innovation is ready to disrupt existing strategic models, the game is already lost and the CEO knows it; but that is a problem for Future Homer, man I don’t envy that guy.

HOUSEKEEPING

I’m working on posts about network effects, a (probable) series on organisational structure and design, team-building and a few others. Let me know if you have favourites or would like me to write about something specific.

poorguard@proton.me