Seth Godin on building remarkable products, brands, and strategy

Executive overview

Most product teams gloss over the four choices that determine everything: who you're for, who you're competing with, who you're trying to please, and how you'll reach people. Each choice shapes the product you build and the life you live.

A brand is a promise — not a logo. Remarkable products are ones people remark about because their lives get better by talking about you.

The core insight: pick your customer before you pick your product, and every other decision follows.

What most product people get wrong

• Empathy is not optional — if users can't figure it out, you made a mistake, they didn't
• Don't run out of time or money; professionals build in buffer for the unexpected
• Without a network effect, you're dependent on platforms you don't control
• Validation: agree in advance whose taste you're matching — the customer's, not your boss's

The four choices that determine your future

1. Choose your customers — their language, budget, problem, temperament, and patience shape the product entirely
2. Choose your competition — the space you operate in sets the rules; compete against Walmart and you'll fight on price
3. Choose your source of validation — align with your boss up front on whose taste matters, then every meeting runs better
4. Choose your distribution — Steam, shareware, mass retail: the channel changes what the product must be

On building a brand

• A brand is what you'd miss if it were gone; a logo is not a brand
• AI will stop being a feature the way electricity stopped being a feature — the question becomes what promise you make
• You can only break a promise once; under-promise and over-deliver beats the reverse
• Loyalty means paying extra to stay; points-based retention is bribery, not loyalty

Being remarkable

• Remarkable means worth making a remark about — not viral gimmickry
• Google's two-button homepage was a promise: "we're smart enough to take you where you want to go"
• Microsoft Word beat WordPerfect through network pressure: users insisted colleagues migrate
• Build so that telling a friend raises the speaker's status — that's the engine of word-of-mouth

Tension and strategy

• Tension is not stress — tension is "it might not work," the gap between a promise and its proof
• Every great strategy has tension at its center: you assert something, the audience wonders if it's true
• Seeing the system means naming the invisible rules that protect incumbents and scare out challengers
• Safety is risky: if something scares you, ask which system is threatened — that's your signal

Taste and standards

• Good taste: knowing what other people want just before they do
• Quality means meeting spec — if you meet spec, ship; if the spec isn't good enough, raise the spec
• High standards = relentlessly improving the spec in service of the people you're working for
• Perfectionism that won't ship is hiding, not quality

On using AI as a writing tool

• Seth used Claude to ask "what did I miss?" on lists — it surfaced items he hadn't considered
• Used it as a patient editor: "what claims here are you not sure I can sustain?"
• Every word published was still his own; AI served as a thinking partner, not a ghostwriter