Some Things I Know (And Some Things I Don’t)

What have I learned from two years of writing?

DALL-E/Every illustration.

Was this newsletter forwarded to you? Sign up to get it in your inbox.


I have been thinking professionally for over two years. I’ve spent roughly 730 days exclusively focused on understanding which technology companies win and why. It has been the privilege of a lifetime. When I started writing Napkin Math, there were 17,000 readers. There are now over 70,000 of you reading what our team at Every is producing. 

Still, while those numbers are gratifying, they aren’t the point. What matters is truth. What matters is insight. Did my two years of effort elicit any of that? To figure it out, I reread everything I’ve written over the last two years. Then I uploaded all of the text into Claude and ChatGPT to summarize my work and pull out insights, and synthesized their outputs into the sections below. 

You can think of this experiment as a text-based mind palace. Each sentence is a carefully considered door in the hallway of that dwelling. You can open each door by clicking the link and reading my essay about that idea. At the end of this experiment, paying Every subscribers will learn what prompts and tools our team at Every uses to do this sort of research. Let me know if you like this idea—I think there is further to dig. 

Two years of work in 924 words

Which companies win and why? 

At the most fundamental level, the winning company is the one whose culture matches the market it is chasing. 

Great companies sell outcomes, not products. 

By matching the mission of the company to the market it serves, sales become dramatically easier

Investors can find these companies by taking a borderline spiritual approach of sensing the fit of the founder to the market

Even if you have an incredible product, if the public narrative is bad, it doesn’t matter

Corporate finance is storytelling, not math. 

The proper way to analyze a company is to recognize that every number is manipulatable under U.S. accounting laws

This is by design and is net beneficial for investors because it is a clear signal about what matters to a management team. 

By understanding the technical underpinnings of each line in its income statement, you know what truly matters to the company. 

What’s the most important competency for a company?

In a fight between quality product and quality distribution, distribution wins.

How you monetize will ultimately determine what you have to build, and whether you need your users to be addicts or missionaries. 

The invisible costs of distribution can sometimes matter more than the legible ones that come with each distribution decision

The point of distribution hacks is to cheaply acquire customers and get network effects going as quickly as possible.

Technology investing is about knowing when to be a lemming and when to be a maverick. 

The venture capital community has completely abandoned its spirit of adventure in favor of lemming behavior, doing whatever is popular with limited partners. 

This leaves a space for investors who are willing to embrace scientific risk and for founders who jump off the VC treadmill

Most surprisingly, there appears to be no correlation between interest rates and large technology companies being built. 

Risk is too often misunderstood in tech and is substituted for heuristics

Boundary-pushing platforms and publishers win. 

Whether you are a creator or a platform, the internet’s attention competition dynamics force all participants to ever-greater levels of moral promiscuity

Consumer addiction is the blood sacrifice required for platforms to grow, which is why every one of them is pursuing an algorithmic TikTok-influenced “For You”-style feed

The only way for platforms to get off that treadmill is to vertically integrate, but no company has successfully pulled that off yet

As an individual actor on the internet’s grand stage, you should recognize that we are still early in understanding its impact on the world, so you should take big creative swings

Ultimately, if you aren’t a platform company, being a media company is very, very hard. 

Devote yourself to the cause of your life. 

While it is tempting to think that the next promotion, job, or possession will fill that hole inside of you, no earthly thing ever can

In tech, too many people try to make themselves feel whole by working at a startup, despite the choice being deeply irrational

Even if you use the best software available, you are never productive enough to be happy. 

You have to live the unoptimized life

You cannot sacrifice your family on the altar of your ambition

Power is tempting but only invites bottom-feeders to suck you dry

True happiness is found in spiritual peace, loving family, and financial security

True fulfillment doesn’t come from having bigger things; it comes from being made more full by simple things

Utility maximization is futile. Devotion to being you is all that matters

Software platforms always win in the long run. 

Software companies win by becoming platforms, not point solutions. 

The best software companies progress beyond a per-seat-sold model to directly tie their revenue to the outcome they are providing

As AI makes creating new software cheaper, it is increasingly challenging for horizontal point solutions competitors to retain market share against full-stack vertical competitors because the former don’t hold the mission-critical data. 

AI makes creation go to zero. 

The internet made it free to send digital goods to users, and AI takes creation costs down to zero

Companies that are able to easily distribute AI tech to users will take the majority of the profit pool

Value will be found in replacing entire trains of thought, or in more corporate terms, automating away entire workflows

AI won’t enable one-person billion-dollar companies, but AI agents will give knowledge workers superpowers to get more done

Crypto’s problems are fundamental to the technology, not the culture surrounding it. 

While there are many promising, nerdy, and weird things (always three positive signals in startups) that tokens can do, this potential cannot be fulfilled. 

Because each token’s utility is compared on the basis of financial return versus social function, the cool stuff crypto can do gets abstracted away, reducing its functionality to a single metric: internal rate of return (IRR). 

If you are selling to a crypto-native audience, a crypto-native product can be a good strategy—there are just a relatively small number of use cases. 

The crypto doom loop of early adopters cycling through various tokens to juice returns is ever-present and unlikely to be solved

Making things cheaper is a moral necessity. 

Technology companies are an ethically and spiritually righteous pursuit because they make goods and services cheaper for everyone

The ability to build something beautiful and share it with the world is one of life’s highest callings

Face computers rule. 

The Vision Pro is a genuinely novel invention that will dramatically improve our relationship with screens and applications by allowing us to be more present. 

How to pull insights out of your own writing

This exercise was one of the more gratifying things I’ve ever done, and AI was key to making it happen. One of AI’s greatest strengths is its ability to pull insights out of large bodies of data. Even if you’re not a writer, doing a version of this with your journals, emails, or other writing can generate incredible understanding. If you want to do it yourself, here’s how: 

Learn more

This post is for
paying subscribers

Subscribe →

Or, login.

Read this next:

Napkin Math

The AI Hardware Dilemma

Why new devices are flopping—and how they might succeed

1 May 2, 2024 by Evan Armstrong

Napkin Math

The Case Against Sam Altman

The board is bad at their job, but they may not be crazy

2 Nov 21, 2023 by Evan Armstrong

Napkin Math

Falling Out of Love With Michael Lewis

The complicated demands of business writing

17 Oct 11, 2023 by Evan Armstrong

Every smart person you know is reading this newsletter

Get one actionable essay a day on AI, tech, and personal development

Subscribe

Already a subscriber? Login