The Age of Incumbents

Distribution + Data Become Concentration Functions

Mar 02, 2024

This is a weekly newsletter about the art and science of building and investing in tech companies. To receive Investing 101 in your inbox each week, subscribe here:

Technology is a versatile little word. It can mean a whole host of things. Some people think technology began with computers, or software, or the internet. Other people think technology didn't really get going until ChatGPT came out.

But the reality is technology is simply "the application of scientific knowledge for practical purposes." So fire, or early stone tools, these are all technology.

However, even if technology as a concept has been a round for the entirety of recorded human history, technology as an asset class is relatively new. When you think about the stock market as one reflection of what "assets" exist you've had railroads, mining, construction, industrials, etc. to reflect exposure to a particular value chain. Technology has always played a part in all of these. But it wasn't until recently that asset managers had to start thinking about exposure to "tech" in a more pure form.

Takeover of Tech

A fascinating exercise that I wish I could take a lot more time unpacking is to look through the history of the S&P 500 as an almost archeological dig of business history. Since its creation in 1957, it has attempted to reflect the overall market by building an index of some of the biggest companies, weighted by market cap.

Again, technology has played a part in basically every company on the list in one way or another. In 1957, there were even some technology bastions included that you'd recognize, like IBM, or Motorola. Folks like HP and Intel got added in 1974 and 1976 respectively. But by and large, the biggest companies were very different than today.

In 1996, for example, the largest companies by market cap were GE ($136B), Royal Dutch Shell ($128B), Coca-Cola ($117B), and ExxonMobil ($102B). Intel and Microsoft got into the list of biggest market caps in 1998, and tech has been an increasingly larger part of that list ever since.

Today, 7 of the top 10 companies by market cap are pure tech companies (not to mention Apple is 44% of Berkshire Hathaway's portfolio; by far their largest position).

While technology as an asset class has become bigger, and bigger, its become pervasive. Everything runs on internet time now. Everything moves much faster. You can look at just how fast the world of interests in particular tech companies turns over. Take this video for example:

But even with that video, when you get to ~2012 you notice how much the change slows down. The biggest platforms stay the biggest. So what happened?

Revolution & Replacement

One aspect of the early internet revolution was that it was just that: a revolution. There were old ways of doing things and now there were new ways of doing things. When you think about the biggest platforms, they were either replacing dinosaurs, or inventing an entirely new thing.

You can paint hundreds of similar pictures:

Tesla building in the blue-ocean space of EVs against 100+ year old auto companies.
When Oracle got started in 1977, it was replacing IBM, who was already 66 years old as a company by then.
In 1999, Salesforce was displacing older companies, like SAP (1972), but was also introducing the world to SaaS, which let it beat out newer entrants like Siebel and Microsoft Dynamics, both launched in 1993.
In 1997, Netflix's biggest competitor (Blockbuster) was only 12 years old, but delivery (and later streaming) were revolutions.
In 1983, Intuit's Quicken was replacing paper accounting and accountants.
Even Uber, founded in 2009, was revolutionizing taxis, which had been around since 1897.

While there are still plenty of startups that are looking to revolutionize certain categories, and displace old incumbents, things have changed.

Dynamic Dinosaurs

The age of lumbering dinosaurs is giving way to an interesting new paradigm: dynamic dinosaurs.

Again, caveat, caveat, caveat. There are plenty of companies that are still slow-moving giants looking to be displaced. But the low-hanging fruit of un-innovated categories and ripe-for-disruption incumbents is becoming slimmer and slimmer.

I hear pitches all the time about next-gen Datadog, next-gen Shopify, next-gen Airbnb, next-gen Procore, next-gen Toast, next-gen Snowflake, next-gen Databricks, next-gen HashiCorp.

I remember investing in Attentive in 2019, and talking about it to someone as the "next-gen Braze." The person I was talking to pushed back: "what are you talking about? Braze is only 8 years old. Braze is the next-gen Braze!"

Does that mean we shouldn't try to replace those companies? Absolutely not. That's the entrepreneur's calling: opportunity canvassing. Looking for weaknesses in markets they can take advantage of. But the reality is that Snowflake or Shopify are nothing like the WordPerfects and IBMs of yester-year.

As playbooks become more well-known, the arbitrage goes away. I've written about this before:

"In 1970 you had a <1% chance of stumbling on someone working in tech. As the industry was just getting started in the mainstream during the 70s there was a total of 450K people working in the industry. Today? 12.2M people. 1.7M of them work for Amazon, Apple, Google, and Facebook alone."

So as startups set out to displace 100+ year old companies, recently crowned incumbents, or peer startups, there are some critical things to consider.

Tasty Addressable Markets (TAM)

Peter Thiel has written a fair bit about markets and competition:

"Competition is overrated. In practice it is quite destructive and should be avoided wherever possible. Much better than fighting for scraps in existing markets is to create and own new ones."

The reason for that is because big markets attract big eyes with big dollar signs in them. The bigger a market is known to be, the more it floods with people interested in getting a taste of it. For example? Business spend. There is over $1 trillion of corporate spend. Big opportunity. And (quite) a few people have noticed.

In 2019, you could have thought the space was pretty well wrapped up. You had big incumbents like Concur (1993) and AmEx (1850), and initial entrants like Expensify (2008) and Brex (2017). Do you need another one? But 2019 was the year Ramp launched. And they've grown to $300M+ of annualized revenue. What's more? There have been some big new entrants into the space, including Stripe, Rippling, and Navan (fka TripActions).

Big markets attract big attention. But big attention often favors incumbents. If Rippling already has a massive install base in enterprises, and is running sensitive systems for them, doesn't that give them an advantage? The advantage is so often driven largely by existing distribution.

Distribution Is King

Now this is one you've probably heard a lot before. But the best example of this is Microsoft. The example? Slack vs. Teams.

Microsoft has done similar launches, like Loop vs. Notion, or Designer vs. Canva. Now... is ANYONE using Loop or Designer? I don't know. But I was one of the people in 2019 saying "who da eff is using Microsoft Teams?" But that's the power of built in distribution. It's just easier to use that vs. a newer detached solution.

Interestingly, distribution can be technologically supported by existing scale (e.g. OpenAI with Microsoft) but sometimes get established in an instant. When ChatGPT came out in late 2022, Yann LeCunn described it as "not particularly innovative," and "nothing revolutionary." And he's not wrong. But the interface was the first time a lot of people had been exposed to generative AI. Shockingly, ChatGPT wasn't a technological revolution, but an awareness revolution.

Seemingly overnight, OpenAI had an IBM-level awareness. So when they're then able to also weave DALL-E 3, a GPT app store, and (hopefully soon) a Sora video model all into one interface? That's powerful. When people go looking for APIs for their enterprise tool, they most often lean on awareness. The “IBM Effect.” And one of the big reasons OpenAI was able to catapult to $2B+ of revenue was because of that awareness.

Data Is Eating Software

Speaking of AI, it introduces the other element of incumbent power reinforcement: data. In the words of Martin Casado:

"Software ate the world, and data is eating software... We have moved to businesses competing on how well they use data and now how well they use software, which is a very big shift."

The first generation of tech incumbents (Amazon, Facebook, Google) was largely driven by eyeballs — that was the hot commodity! They each created unique services that attracted eyeballs, and then they were able to monetize that to the tune of hundreds of billions of dollars.

The next generation asset will be data. The more data you have, the more you'll have people lining up for your product. Already, we're seeing eyeballs turn away from attention-seeking websites, like StackOverflow, and towards AI services, like code completion.

Now, you'll see platforms with data, like Reddit, StackOverflow, Tumblr, etc. look to monetize that asset. Granted, the dichotomy is different. Facebook and Google had the eyeballs and made money on them. Now Google wants MORE data, and is willing to pay for it.

But the reality is the data they're paying for is supplementary. Someone like Google can pay $60M to Reddit because they already have so much monetized data. But the incumbents are still positioned to make dramatically more money from those data-driven assets than the data providers. And the combination of access to data + massive install bases are the perfect combination to keep incumbents deeply entrenched.

So what could possible stop the concentration in tech from staying concentrated?

The Canaries In The Coal Mine

There are plenty of reasons not to loose hope when it comes to displacing incumbents. My focus, instead, is on articulating why its important to not just wave off the threats. The way that things are evolving, it will be increasingly harder to unseat established players. But that doesn’t mean there won’t be cracks in their armor.

Deglobalization

One of the biggest potential opportunities will likely come from potential Black Swan events as the broader world changes. The biggest potential future threat is global conflict. Palmer Luckey has made this point about the dependency companies like Apple have on China:

"You look at situations like US Congress passing the chips act. $50 billion. And they call it a once in a generation incredible subsidy to the semiconductor industry. $50 billion is a lot of money, but you know what's more money? The $250 billion that Apple had to pledge to invest in Chinese education and manufacturing just to renew their independent operation deal in China for some unknown number of years. You have one company investing five times what the United States is in that type of advanced manufacturing."

These massive investments bring with it massive inter-coupled risks for incumbents like Apple (and they're not the only ones). Palmer goes on to make the same point about the risks:

"Companies like Apple have deep ties to China. I don’t know if I would say inappropriate, so much as existentially crippling. Let's talk about a really big problem that's an elephant in the room that doesn't get talked about nearly enough. The analysts, the financial weenies, they look at these companies and they try to come up with what the risks are. ‘Oh man, rising interest rates are going to hurt their sales. Could it be that they're not able to capitalize this plant that's over here,’ but people don't like to talk about the real risk. The existential risk, the risk where Taiwan is taken by China and they either fully deprive us of advanced capabilities that we would certainly deprive them of in any kind of wartime scenario or alternatively that they unilaterally act to cripple our largest companies. Think how much trouble Apple would be in if China decided they wanted Apple to be in trouble. Almost all their manufacturing is there, they cannot exist as a company without that. You have a nearly three trillion dollar company, the biggest powerhouse of economic value in the entire United States totally beholden to the whims of one person in China. That's crazy, that's unprecedented, but people don't talk about it including Apple, because the alternative if it happens it's too unthinkable. It's a risk you can't even talk about, you can't even discuss this risk rationally because it's so extraordinarily bad of an outcome."

Another incredible dependency is in semiconductors. Everyone is talking about how Taiwan is Arrakis; the most important asset in the universe. Why? TSMC. Not just Nvidia, but 90% of the chips made by Nvidia, Apple, and Broadcom are made by TSMC. Those aren't casual displacement opportunities — those are risks with global ramifications, or, in the words of Palmer Luckey, “existential risks.”

Culture Wars

Global conflict has the sharpest teeth, but culture has the most powerful ideological teeth. And ideology can be the death of many an exceptional business. The most recent case? Google. After the fiasco that was Gemini, Ben Thompson dug in. After describing Microsoft's "kiss with death" in 2013 over cultural issues, he explained Google's position this way:

"Google is not in nearly as bad of shape as Microsoft was when it held that funeral. The company’s revenue and profits are as high as ever, and the release of Gemini 1.5 in particular demonstrated how well-placed the company is for the AI era: the company not only has leading research, it also has unmatched infrastructure that enables entirely new and valuable use cases. That, though, makes the Gemini fiasco all the more notable."

The ideological divide will continue to sift companies, and depending on the outcomes, it will position or deposition certain companies from succeeding. This isn't just Gemini's cultural weirdness, but its also playing out in proprietary vs. open source AI, labor conditions, hiring, remote work, and a host of other issues.

My perspective is the more we get embroiled in culture wars, the worse everything is going to get. But I can't stop people from getting offended on the internet, and neither can your business. So being aware of which side of the culture war you want to bet on is a critical business decision you have to make. So make it wisely.

Opportunity Canvassing

I'll conclude with this: there is ALWAYS an opportunity to build a massively successful business. Like I said before; the age-old mission of any entrepreneur is to identify opportunities and take advantage of them. All of these canaries represent potential weaknesses to incumbents, but you're not limited to just waiting for big companies to get cancelled, or for the kick-off of World War 3.

One of the biggest opportunities seems to be in-line with a time-tested recipe. Looking for dinosaur incumbents that are far from dynamic, but who represent massive pools of revenue ripe for the taking. Understanding why a category has remained under-innovated is more critical than ever because if it was easy, it would have been reinvented already.

One example? There's expected to be $741 billion of revenue in call centers by 2030. Massive BPOs of people answering phones, etc. Recently, Klarna shared a data point that ~66% of their customer support tickets could be handled through AI. A job done today by 700+ people. Does that mean this is a good pool of revenue to go after? Or a bad one? Just because AI can handle things, doesn't mean the job is done. There are workflows, support, model training, customization, etc. that need to go with it.

But candidly for some of these buckets? I'm most excited about Accenture's ability to help people try and roll AI out. They're already investing $3 billion in their data and AI practice.

Other buckets? Manufacturing in the US represents $7 trillion. With the headwind of deglobalization, how much of that will be up for grabs? Whether its robotics, new-age machine shops, processing systems, etc.

The opportunities are out there. But here's some key takeaways:

For a long time people wrote off questions like, "why won't Microsoft / Facebook / Google just do this?" But that's not a great thing to ignore anymore. Those are legitimate threats, and they’ve never been better positioned.
It's not enough to say "this is better than XYZ Incumbent because the user experience is better." We're not replacing legacy dinosaurs; the dinosaurs today are much more dynamic.

Happy hunting!

Thanks for reading! Subscribe here to receive Investing 101 in your inbox each week:

Investing 101