Business strategy

Frank Niu on finding customers before you build

Frank Niu joins How Does This Get Big? to explain how his audience shaped his startups, and why trust and reliable software matter together.

13 October 20262 min read5 min listen

Listen to the audio analysis

Create your own

Written for builders and creators.

The traditional moat for software engineers—the ability to write complex, elegant code—is being hollowed out by AI. And honestly, a much more aggressive kind of competitive advantage is taking its place.

Editor's note

Why this matters now

Frank Niu, a former Netflix engineer, joins Andrew Meng and Thomas Kim to talk about the part of a startup that engineering skill alone leaves unresolved: reaching customers.

His own products gave him different answers. A language-learning app struggled to reach people consistently. For a later trading-related product, he worked from an audience he already knew.

Listen if you're choosing what to build next. Niu's useful challenge is to explain how you'll reach the buyer with the same care you give the product itself.

The source

What it says

Distilled from the original. The notes above and below are the editor's own.

Engineering skill leaves a customer question

Frank Niu, a former Netflix engineer, discusses his move into content and startups with Andrew Meng and Thomas Kim.

He argues that the ability to build software needs to be paired with a way to reach customers. His account draws on products with different audiences and different consequences when something goes wrong.

A product needs a repeatable way to reach people

Niu describes a language-learning app that struggled despite the work put into its features. Reaching users consistently remained a problem.

For a later trading-related product called Mirror, he worked from the interests of his existing audience. He says his content helped attract early demand.

The useful distinction is how each product found its customers. The interview's revenue and sales figures are Niu's own account, with no independent audit presented here.

Building an audience takes practice

Niu says he once posted as many as 24 videos a day during an early period of content creation.

That is his account of the work involved, with no recommendation that every founder should copy that schedule. He describes finding topics he could discuss naturally and producing content consistently.

Over time, an audience can learn what a creator knows and decide whether to trust their recommendations.

Match the engineering to the consequences

Niu distinguishes experimenting with a consumer product from building something that handles people's money.

For the latter, reliability and regulatory responsibilities become central to the work. A product's popularity cannot make those obligations disappear.

He also discusses using AI coding tools while retaining human technical judgment. Those observations describe his experience and expectations, rather than a measured forecast of jobs or engineering tasks.

Work backward from the buyer

For a builder deciding what to make next, the episode gives a practical question to answer: where does a potential customer already pay attention, and what would make your offer relevant there?

The answer can guide an early experiment in demand while you continue checking whether the product can deliver what it promises.

Editor's note

What to do with this

Write down where your first customers will hear about the product and why they will trust the offer. Then test whether you can reach those people before expanding the build.

Niu also stresses that a product handling money needs dependable engineering. A promising audience gives you a place to test demand. It leaves the responsibility for safety and reliability with the team. For audience-building choices, continue with Meng and Kim's personal-brand discussion.

Find more from the show in our startup-growth playlist.

The original

How This Retired Big Tech SWE Built an App That Made $30k in 2 Weeks

How Does This Get Big? · 6 October 2026