The Strategy That Shapes The Tech World

"Commoditize your complements" shaped everything from early computers to electric cars and may shape the AI future.

Photo by Jen Gunter on Unsplash

As someone who majored in Economics, the nuances of microeconomics have always fascinated me. It’s one of the reasons why I write about business strategy so much, it’s very interesting! Technology companies have used a multitude of microeconomics informed strategies to stay ahead of their competition, and continue to do so to this day.

Today we shall look at one of those strategies; commoditizing your complements. It’s the reason why Google, Tesla, Apple, Microsoft and others are the modern juggernauts that they are. As AI becomes the new paradigm with which businesses hope to shape the future, they are using this very strategy to pull away from the pack.

In economics, a substitute refers to a good or service that can be used in place of another. For example, apples are a substitute for oranges. Consumers may choose one good over its substitute because of preference, price et cetera. If I have an orange farm, and the price of apples goes up, people will want more oranges and my sales will go up.

A complement, on the other hand, is a good or service that is bought together with another good or service. For example, flight tickets and hotels are complements. Cars and gas are complements. Computers and operating systems. Electric cars and charging stations. To name a few. When flight prices to a certain destination go down, hotel owners see more guests. The general framework is:

All else being equal, demand for a product increases when the prices of its complements decrease.

Over the years, technology companies have used this framework to great effect to entrench their positions and maintain dominance. How? By ensuring the price of their complements is as low as possible. The lowest theoretically sustainable price would be the “commodity price” — the price that arises when you have multiple competitors offering indistinguishable goods. Therefore, as a smart company, the goal here is to commoditize your complements.

Joel Spolsky, a software engineer and co-founder of Stack Overflow published a now legendary blog post in 2002 that explained this very topic. He discussed it within the lens of tech companies strategic decisions in the 80s and 90s. As Joel puts it:

When IBM designed the PC architecture, they used off-the-shelf parts instead of custom parts, and they carefully documented the interfaces between the parts in the (revolutionary) IBM-PC Technical Reference Manual. Why? So that other manufacturers could join the party. As long as you match the interface, you can be used in PCs. IBM’s goal was to commoditize the add-in market, which is a complement of the PC market, and they did this quite successfully. Within a short time scrillions of companies sprung up offering memory cards, hard drives, graphics cards, printers, etc. Cheap add-ins meant more demand for PCs.

When IBM licensed the operating system PC-DOS from Microsoft, Microsoft was very careful not to sell an exclusive license. This made it possible for Microsoft to license the same thing to Compaq and the other hundreds of OEMs who had legally cloned the IBM PC using IBM’s own documentation. Microsoft’s goal was to commoditize the PC market. Very soon the PC itself was basically a commodity, with ever decreasing prices, consistently increasing power, and fierce margins that make it extremely hard to make a profit. The low prices, of course, increase demand. Increased demand for PCs meant increased demand for their complement, MS-DOS. All else being equal, the greater the demand for a product, the more money it makes for you. And that’s why Bill Gates can buy Sweden and you can’t.

Since then, the strategy has been used to devastating effect in multiple different facets of the tech industry. From mobile, to personal computing to electric vehicles.

I’d argue the entire mobile landscape today is shaped by Apple and Google’s implementation of commoditize your complements.

Apple

Source: Apple

When Apple launched the iPhone in 2007, they also launched the App Store. “There’s an app for that” was not only a slogan, but an indicator of commoditization strategy. Apple made it easy for everyone to create and distribute apps, by offering software development kits (SDKs) to developers and hosting apps on their App Store that had millions of customers frequenting it. Listing an app on the App Store was also free, although they took a 30% cut of all in-app purchases.

Apps soon flooded the App Store. They were rudimentary and playful at first, but soon became capable and complex. This is because the app market became fiercely competitive as a result of constant iteration and an access to quick feedback, which Apple enabled. Soon, apps made using an iPhone a unique experience that just was not available on a Blackberry or Nokia. This became the iPhone’s differentiator.

Apple could see before everybody else that apps were going to be complements to smart phones. Thus by commoditizing apps they sky-rocketed the demand for their own phones.

Source: Counterpoint Research

The strategy worked, and to this day Apple has sold billions of iPhones worldwide. Interestingly, however, Apple’s once rosy and symbiotic relationship with app developers has soured. Developers are now asking for Apple to tone down on the strict App Store requirements that once worked for everybody, but now only seem to benefit Apple. I wrote about this changing dynamic here.

---

---

---

Google

Google approached the mobile market with a different implementation of the same strategy. When the search giant saw Apple’s App Store announcement, they knew they had to compete. They did this by developing the Android Operating System and open sourced the handset version of it at no cost to mobile phone makers. Today, Android is the leading mobile OS on the planet, so it’s easy to forget how handsets operated before its existence. You see, handset makers once had to develop their own operating systems as well as manufacture and develop the phones.

Making successful consumer hardware is already hard enough, and having to make great software to go with it turns the difficulty level up to impossible. Only a handful of companies have ever been able to do this successfully. However, Android OS changed this and made mobile phone making easier and cheaper. Companies only had to focus on the hardware and then license Android for free. Today, there are hundreds of Android handset makers globally outcompeting each other every year. The handset market has commoditized.

What’s a complement to handsets? Google Apps. Search, Gmail, Youtube and Google Pay all come pre-set into Android handsets and thus their mass use has exploded Google's revenue over the past 15 years. Google could’ve attempted to compete with Apple and offer their own hardware + software experience in a phone, but they didn’t. They knew Apple were better than them at that. Instead, they went for the smarter option; commoditize the handset market and have billions of users globally use Google Services.

Source: Statista

Tesla

The method to Tesla’s madness has always been fascinating. The company, which started out by making fully electric vehicles, has taken a multi-pronged approach to commoditizing their complements that we must address here.

As I mentioned earlier, it is extremely difficult to make successful consumer hardware products. Whether it’s phones, computers, watches, cars, cameras, headphones. When a company has a successful product, they often fight tooth and nail to keep their secret sauce private. Why? That’s their differentiator. Their patent, or schematics or architecture is what makes them win, and thus they protect it with their lives.

Well, this is true for every company but Tesla. In 2014, Tesla made their patents public. The company explained that despite initially patenting their technology to prevent large automakers from copying their cars, they realized that their actual competitors weren’t just other electric cars, but rather every gasoline car on the road. CEO Elon Musk explained it here:

Given that annual new vehicle production is approaching 100 million per year and the global fleet is approximately 2 billion cars, it is impossible for Tesla to build electric cars fast enough to address the carbon crisis. By the same token, it means the market is enormous. Our true competition is not the small trickle of non-Tesla electric cars being produced, but rather the enormous flood of gasoline cars pouring out of the world’s factories every day.

This was a stroke of genius. You see, Tesla was eventually going to get into the business of providing batteries to other automakers, which they do now. By open sourcing their patents, they began the process of commoditizing the electric car market. Today, the electric vehicle space is fiercely competitive; new entrants are often sprouting offering more range, better efficiency and lower prices. Even Tesla had to lower their new car prices early this year. Lowering the barrier to entry for EV manufacturing results in more customers to sell batteries to.

Furthermore, the company began flooding their largest markets with superchargers - locations where people could charge their electric cars, regardless of brand. By growing the supercharger network, Tesla increases demand for electric cars and opens two avenues by which they can benefit. The company benefits if a new car bought is a Tesla or if it’s a car that Tesla supplied with batteries.

Tesla 2023 Income Statement. Notice the increasing Energy Generation and Services Revenue. Source: sankeyart

---

How Commoditize Your Complements Is Shaping AI

If you’ve been paying any attention to the AI landscape, you’ll notice that the big tech giants such as Meta, Google, Apple and Microsoft are just as involved, if not more involved, than startups like OpenAI and Anthropic. This is uncharacteristic for the tech industry, which is synonymous with new players shaking up the status quo. I have written about this discrepancy here, and most recently here. I’d actually recommend you read those before we go ahead, but if you don’t have the time, let’s power along and you can catch up with them later.

One way that tech incumbents have been making waves in AI is by releasing their own open source large language models. Last year, Meta released their model Llama 2, which is free to use for all developers. To many, this came as a surprise. You see, it costs millions of dollars in computing power and energy costs to train and fine-tune these models. Conventional wisdom suggests you ought to seek a return on that investment. In addition, other players, such as OpenAI, keep their models very secret, never releasing the weights or training data. So what was Meta playing at?

You guessed it; commoditizing their complements.

Meta’s golden goose is ad revenue. They obtain this ad revenue by allowing companies to advertise on their apps. Companies advertise on Meta apps because that’s where users are. Users are there because Meta apps are a great source of user generated content. Facebook updates, Instagram pictures and reels etc are all user generated content. I’ve written about the relentlessness of the content economy here.

Large language models make content generation easier. I could probably ask ChatGPT to write this essay. The result wouldn’t be as good as this, obviously, but it would be ready in a matter of seconds. As these models get better, content will become easier to make and the quality of said content will get higher. As a result, it is in Meta’s best interests to make generation of the best quality content as easy and cheap as possible, thus they are determined to commoditize the large language model.

Imagine if every random tech reviewer could produce videos as highly polished as MKBHD? That would mean more users to watch those videos, and more ad revenue for Meta. This is what Meta will be hoping, although one has to worry about the saturation of the attention economy and how much more content can really scale. I’ll write about this soon.

This isn't unique to Meta either. Last week, data management provider Databricks released their own open source model. Their strategy? More LLMs means more data repositories required which means data management demand goes up. Even Apple of all people released an AI research paper on their small language model. Their strategy is less clear, but I’m sure it will come to light soon.

All these companies are attempting to commoditize the large language model in order to increase demand for their own product or service, which we may not know yet or may be many years away.

Strategy can be stealthy, unexpected and unique. But in order to understand motive, always remember; the end goal is always to sell more oranges.

---

Thanks for reading another edition of Always Learning. If you enjoyed this post, you might like: