Skip to content

What I’ve Learned Searching for a New Startup Idea, Part 2

Tags: startups • Categories: Learning

Table of Contents

(this is part two of a series of blog posts about what I’m learning as I hunt for a new startup idea to work on with my cofounder)

Find the Best Talent

Hire only great people. Eliminate toil incrementally, even at the early stages. Slow growth to avoid hiring B players; this will result in a net decrease in product velocity.

This model breaks at some point in a company’s growth, but I think you can keep a small-and-mighty team for longer than one would expect. This was a key advantage Stripe had in its early days.

Be Curious and Start Small

Have childlike curiosity. Don’t be afraid to build something silly, small, or seemingly useless. Too much filtering of ideas on the outset can cause you to miss something big. All important companies look small and stupid in the beginning.

In other words, a love for the craft, excitement in getting something out into the world, and a desire to change your customers’ lives for the better are good motivators. These aren’t dependent on your calculus of what might be successful.

However, you have to have taste and an intiuituive sense of what is going to work. There’s no formula for this and this is where the risk and personal conviction comes in.

Chatting with Experts is a Superpower

Talking to experts is the best way to learn new domains quickly, and pretty much anyone will talk to you. Be generous with others and they’ll be generous to you. One of the most amazing things about tech is the willingness for basically anyone to help you: people know a rising tide lifts all boats and technology is not a zero sum game.

Some Big Companies Can Crush You

Most big companies are too slow and have too many big opportunities to build something competitive with you, especially if it risks destroying their existing business. Even Intel (TSMC), Adobe (Figma), SalesForce (CPQ), etc can’t figure out to build a startup within their orgs to solve massive needs.

However, some can.

Microsoft. They compete with Slack, Zoom, Notion, JetBrains, SourceGraph, AWS, etc.

Stripe competes with FirstBase, payment link businesses, Recurly, Signifyd, etc.

OpenAI, although growing fast, has continued to have amazing product velocity. I wouldn’t bet on them not doing something, even in the more consumer-y side of AI.

Be aware of what type of incumbent you are competing against and if what you are building could be in their top 5. It doesn’t mean you shouldn’t pursue it, but plug this into the overall calculus.

Information Architecture Hugely Impacts Product Velocity

How you communicate, both internally and externally, can massively impact your velocity. Everyone hates meetings, yet most companies are filled with them. It’s the result of poor discipline around information architecture.

This is something you can design and think about apriori. It’s much easier to establish information architecture decisions—which are largely cultural—before you have a larger group of employees.

Big Ideas Require Focused Building and Spontaneous Collisions

It’s "easy" to find an idea for a sub 10M ARR business. Attempting to find an idea that:

  • Has good founder-fit, GTM fit, and PMF fit
  • Is a customer you want to work with (subset of founder fit, in a way)
  • Is not speculative (i.e. will take 5-10 years to know if the core technology works)
  • Is big (> 100M ARR)

Requires quite a bit of luck and is not a deterministic process. However, I you can stack the deck in your favor.

Talking with experts in the field, understanding their pain without leading the witness (i.e. Mom Test), doing analytical research (market sizing, trends, etc), ghost product testing (splash page, cold sales, etc) can get you a lot of data before you touch any code. It can help both eliminate bad ideas (like building a natural language to SQL product that no one wanted) and help you become a product expert quickly.

Then, working closely with passionate beta customers and building an a small, sharp MVP quickly is the best way to incrementally stumble upon a great idea.

Big Ideas Are Found on the Frontier

The germ of many big businesses (GitHub, SnowFlake, Jasper, Figma) was finding a new peice of technology that has not yet been applied to a domain you are familiar with.

Spending time on the technology frontier is a great use of time. Reading papers, side projects that intentionally explore something new, spending time with interesting & curious people. These are all ways to encounter a new idea which could be the germ of the next business.

Make a Bet on a Trend

Investors call this a "why now moment". Another way to think about it is a catalyst: some external event that is generating a compelling reason why now is the moment to build something new.

It’s easier to build a big business when, at the core of it, there is a contrarian (or, at least, non-obvious) bet on the way the world will change over time. A trend that is small now but will be huge later.

Do the work to find a trend or opportunity that you feel excited about betting big on.


I’m a big believer that you can’t do two big things at once. Once you are willing to make a bet on a big idea, you need to go all in and quit everything else.

This is especially true if you have kids (like I do). Business is not the most important thing in our lives, even if it takes up the most quantity time. You need to be ruthless with eliminating time sinks and outsourcing areas of our lives that are not the highest and best use of our time to be able to maximize focus on a business.

Consulting Can Be Useful to Find New Ideas and Future Customers

However, you need to very careful to:

  • Not get trapped in a ‘pit of success’. Money should not be the motivator for consulting.
  • Optimize for jobs that introduce you to a new industry or technology.
  • Establish clear expectations: no long term contract, primary goal is to build another business, must be able to leave at any time, cannot be full-time.
  • Ensure the buyer is highly motivated for you to succeed in your contribution. Make sure you are building something to critical to the buyer and they will work to help you learn and transfer necessary domain knowledge.

Many businesses were formed out of seeing a massive problem in an industry they were doing consulting in. I’ve found consulting to be useful is getting a deep understanding of a deep domain.

Raise Money when You Need it

Money raised as a metric of success always struct me as strange. Raising money is a promise to do something great for someone who just put their hard-earned savings into a risky venture.

The mimetic urge as a second-time founder is to raise money right away. Instead, I’ve really felt like it’s the right move to slowly + carefully find a deeply exciting idea, prototype + build the initial product, and thoughtfully discern what sort of capital support the business needs. We hope to build something venture scale, but it’s possible that we run into a deeply exciting idea that risk capital isn’t the right fit for.

As a friend says, "a business can’t escape it’s capital structure". Making sure the business the business never needs to escape the incentives created by its capital stack is critical.