Thoughts on Value Chains, and Why You Really Need To Get a 'Grip'
European Straits | Work in Progress
Hi, it’s Nicolas from The Family. Here are some thoughts expanding on my recent essay on capitalism, for all of you who are interested in corporate strategy.
Agnes Callard, a professor at the University of Chicago, recently expressed interest in my expanding on the idea of a ‘grip’—“the control over multiple parts of the value chain that eg Amazon has, but Uber lacks”. And so here’s a first draft. Since I’m still refining my idea and want to expand it further, notably by adding illustrations, it’s still part of the “Work in Progress” section of European Straits🤗
1/ In my latest piece on capitalism, I explained that there are two keys to succeeding as a capitalist enterprise: you need to pursue increasing returns to scale, and you need a ‘grip’ on multiple sections of your industry’s value chain. Here’s the related quote, after which I’ll expand a bit on the idea:
The lesson I draw from [McDonald’s] is that of the ‘grip’: every successful enterprise in the history of capitalism has focused on relentlessly pursuing increasing returns to scale, which requires focusing on certain parts of the business, while still maintaining a grip over the rest of their value chain. And how do you get that ‘grip’? By operating at least two separate links in the value chain.
For McDonald’s, it’s the ‘system’ and real estate. For General Motors, it’s the brands and assembling the cars. For a publishing house, it’s editing manuscripts and distributing print copies at a large scale. Wherever you find capitalist success, you’ll find these two things: increasing returns to scale, and the ‘grip’.
2/ First, what’s with this idea of an industry value chain? Aren’t we supposed to be in a world of platforms rather than pipelines, as once explained by Marshall W. Van Alstyne, Geoffrey G. Parker and Sangeet Paul Choudary? I shared quite a lot on the idea of an industry value chain in my 2016 essay The Fives Stages of Denial:
An industry is that whole set of businesses that work together along an industry-level value chain. They’re like gears that fit together perfectly, linked by a common culture and practices developed and optimized over the course of decades.
Each gear on the chain is also related to its own sector: other businesses that perform the same activities and split up the market amongst themselves (enterprise market upstream, consumer market downstream).
3/ Let’s have another attempt at defining an industry value chain: it’s all the links that separate a job to be done (at the bottom) and the raw resources that you need as an input to do that job (at the top). The value chain is everything that happens in between: raw resources need to be transformed into a finished product, which in turn needs to be moved closer to the end users, and then you need to sell it (and then you might have to provide customer support, depending on the nature of the product).
4/ The top-to-bottom analysis of a value chain suggests it’s a top-to-bottom sequence. But it actually depends on what business you’re in:
If the customer needs a good to do the job, then the process effectively happens from top to bottom: you need to extract the coal so as to manufacture steel, which in turn will be turned into car parts, which in turn will be assembled into a car, and then that car will be moved to the car dealership, where it will be sold to a customer for whom it hopefully does the job.
If the customer hires a service to do the job, then it will likely happen in the reverse order: the customer needs that parcel delivered, so you’ll get the package from them, load it on a small truck, which will bring it to the airport, then when the plane has landed closer to the end destination, another truck will be used to deliver the parcel to its recipient, and the job is finally done. In the case of a service, the sequence is bottom-to-top rather than top-to-bottom.
5/ An industry’s value chain is in a permanent state of imbalance. The different links (which are effectively different sectors: manufacturing car parts, assembling, selling cars, etc.) are in a constant struggle with one another to try and grab a higher proportion of the total value added in the entire chain. Here are some examples, from the music industry:
When selling music online started to become a thing, the strongest link in the value chain was the record labels. They decided on hostility toward new entrants: they sued Napster, they refused to license rights to other music startups. In the end, the record labels were so strong that they managed to strangle the entire value chain, clearing the field at the bottom: nobody was (legally) distributing music online anymore—it was all about piracy.
This lack of viable competition made it easy for Apple to counterpunch. As told by Walter Isaacson in his biography of Steve Jobs, the record labels were so desperate in the face of piracy that Jobs had no difficulty imposing on them what ultimately triggered iTunes’ ascent: the ability to sell music by the song rather than by the album. Suddenly Apple gained the upper hand in the industry, extracting more value from the record labels up the chain.
This is when record executives realized that they had gone too far in squashing every startup in sight and that they needed to diversify their distribution channels, so as not to depend on Apple when it came to selling music online. They decided to deal with Apple’s competition and started negotiating deals with the likes of YouTube and then later Spotify.
For instance, in the car industry, selling cars comes with diminishing returns to scale: it’s hard to scale up a car dealership. On the other hand, assembling cars comes with increasing returns to scale, because it’s easy to increase productivity (the old combination of scientific management + the American system of manufacturing). I’d say that manufacturing parts is somewhere in-between: more scalable than selling cars, but not as scalable as assembling them.
As for the music industry (to echo the example above), there were increasing returns in licensing rights, marketing records, and moving them around using dedicated logistics. On the other hand, two sectors clearly featured diminishing returns to scale: one was selling records to the public (the biggest you could become was Tower Records); the other was actually recording the music (which is why artists in general have a hard time claiming a large share of the value added).
Are you starting to see a pattern here? Most large, successful capitalist enterprises are those that focus on sectors where you can generate increasing returns to scale (car assembly lines, record labels). On the other hand, those who belong to sectors in which you can’t scale...don’t scale. As a result, they can’t defend their share of the value added and are condemned to live with ever-diminishing margins.
7/ And then there’s this additional idea of the ‘grip’. It’s not enough to dominate one sector with increasing returns to scale. For instance, you can’t just assemble the cars without also owning and marketing the brands. And you can’t just own the rights to recorded music without also marketing artists to the public. Why is that? Why do you need to be present at two different levels?
One reason is that you don’t want a foe with an equal ability to scale off growing somewhere at another level of the value chain—otherwise there’s a probability that you will lose the battle for capturing the value added and have to trim your margins as a result. This is what has been happening to Spotify in the music industry: they certainly have increasing returns to scale, but not as much as the record labels, and so the record labels win when it comes to grabbing shares of the total value added. It’s not enough to enjoy increasing returns to scale, you need your returns to increase more than at any other level of the value chain. That’s how you end up dominating an industry and capturing most of the value!
8/ Another reason echoes my idea of the Southern Side and the Northern Side. It’s not enough to enjoy increasing returns to scale if your market position can be wiped out in an instant because of a self-inflicted wound (like Uber in 2016) or if the context changes (as it did for Facebook following Trump’s election). Because you never know what will happen to you if things get shaky, it’s much more secure to have a foothold at two different levels of the value chain:
It’s about balance. You need at least two legs to keep your balance. Since we humans have two legs, we can stay on our feet no matter what—like if we’re on the train and it suddenly brakes, or someone bumps into us on the street, or if we’re a boxer and get hit, we can adjust quickly and stay standing. On the other hand, someone with only one leg needs crutches; and if they don’t have them, they’re unable to stand for very long. The imagery might be a bit far-fetched, but I’m pretty sure it’s the same for a business aiming at dominating its value chain.
It’s also about defensibility. Software (as in “software is eating the world”) comes with nice network effects that generate increasing returns. Alas, because software depends heavily on the users, returns to scale can disappear in an instant if something bad happens. Unlike the increasing returns that existed in manufacturing, software-induced increasing returns are precarious. It’s difficult to retain them in the midst of a crisis, because you don’t own your users like you would own a factory. When those users are gone, they’re gone! Just ask MySpace or Yahoo.
This explains the Netflix strategy. When they were all about streaming, they were only present at one level of the value chain. It came with increasing returns, but they were still losing the fight to share in the value added with the big rights holders up the chain. This is why they decided to expand into producing original content. The returns aren’t as high, but it makes it possible to have that ‘grip’.
Whatever becomes of the entertainment industry, Netflix is walking on two legs now: like us humans, they can adjust quickly and stay standing. Plus, all that original content is an asset that nobody can take from them—unlike users of their streaming platform, who can decide to cancel their subscriptions and flee to Amazon, Hulu, Disney, or HBO in an instant.
9/ Another idea is the diversity of paths that successful companies follow to arrive at a solid grip. There are four interesting cases that I’d like to mention:
Ford Motor Company. It started with a strategy of vertical integration: Henry Ford wanted to do everything in the nascent value chain that was the car industry, from parts manufacturing to selling cars to the public. It made it possible for him to beat the competition from a productivity standpoint (hence the Model T that you could have in any color, so long as it was black). But then General Motors imposed the model of a car company focusing only on assembling the cars and marketing them, while all the rest of the work (in particular manufacturing parts and dealing with customers) was abandoned to weaker, independent links in the value chain.
John D. Rockefeller’s Standard Oil. You might remember it as a vertically integrated trust, but it wasn’t always like that. At first, Rockefeller was focusing on transporting and refining petroleum. Then he started to sell gasoline to the general public so as to strengthen its brand and retain the margins at the bottom. Then he expanded into oil exploration, not because it came with increasing returns (it really didn’t) but because he needed to tighten his grip on the whole value chain and, above all, to secure assets whose value was going up by the day—and he feared dependency on a third party that could perhaps one day own all the oil fields!
Berkshire Hathaway. This may seem different because it’s in the financial services industry. But there’s still a value chain—with raw resources (capital) at the top, and a job to be done (funding businesses and households) at the bottom. Berkshire Hathaway is interesting because over time it has built a very special ‘grip’ that contributes to generating exceptional returns: down the value chain, they have the Warren Buffett brand (the folksy aphorisms, the annual letter, the advising Barack Obama, and so on); at the top, they have access to more, cheaper capital thanks to Berkshire Hathaway’s legendary “insurance float” (read Note #7 in my 11 Notes on Berkshire Hathaway if you want to know more).
McDonald’s. This is best explained by this great scene of the movie The Founder 👇 Ray Kroc (the “founder”) had the brand and the ‘Speedee System’, but that didn’t translate into a capacity to expand faster: the brand was simply not valuable enough at the time. This is when Harry J. Sonneborn, who went on to become Kroc’s CFO, suggested buying land for two reasons that are well explained in the scene: you get ahold of assets that you can then use to raise capital (that’s increasing returns through the detour of raising more money); and you get a ‘grip’ over the franchisees, which makes it easier to keep them in line and retain most of the value added.
10/ I’m still not sure this kind of thinking is widespread within the tech industry. I suspect the reason is that the concept of an industry value chain sounds very industrial/20th-century and we’re all blinded by these discussions that “everything is different now” and that it’s all about “ecosystems”,“platforms”, “APIs”, “aggregation”, “software-as-a-service”, even “micro-services”.
I agree that those are all important concepts, but I stick to a few very important ideas:
Value chains are here to stay if you realize they’re all about connecting raw resources at the top and a job to be done at the bottom. The boundaries between them might be blurred by software eating the world, tech companies turning into giant conglomerates, and multi-sided businesses becoming the norm. But you need to see beyond the hype and the complexity and realize there is still such a thing as value chain analysis.
A value chain is itself a complex system that one enterprise can dominate only by controlling a few strategic points. Controlling one strategic point is good, controlling two is better! It’s like acupuncture: to fix pain in the back, you need to place a needle in various different parts of the body (for example, a needle in the right ear and another in the forehead).
All in all, any business that wants to do capitalism successfully needs to focus on increasing returns; yet the obsession with pursuing those shouldn’t come at the expense of getting a ‘grip’ over the rest of the value chain. At some point you need to integrate that business line even it’s not as scalable (Apple Stores, Netflix’s original content, Airbnb’s building their own hotels) if it can help you tighten that grip over the value chain—and because, as Nivi once wrote,
The best products require unique means of scaling. The delivery of the best products is tied into the product itself. For example, look at Apple’s efforts to develop new manufacturing techniques and stores for its products.
What do you think? Do you have other interesting examples of businesses that work like this—including ones that I could dig into as part of my “11 Notes” series? Also, I realize that some graphics could be useful to explain the concept of the ‘grip’, but I’m still struggling with what kind of tool I should use to turn my sketches into neat graphs with circles and arrows. Let me know if you have something in mind!
From Normandy, France 🇫🇷