Growth tactics, retention strategies, and writing for novelty with Julian Shapiro

Executive overview

Most startups rely on paid ads or SEO to grow — channels that are expensive, saturated, and volatile. Product-led acquisition (PLA) flips this: the product itself recruits new users through normal use, at zero marginal cost. Retention follows a parallel logic: users who build non-transferable state inside a product are too invested to leave.

Writing ties it together — novelty is what earns attention, and without a deliberate creative process, most writers quit before good ideas arrive.

The compounding advantage goes to products that make users richer the longer they stay.

Product-led acquisition (PLA)

• PLA means the use of the product grows the product — no ad spend required.
• Four categories: debt settlement, conversation access, billboarding, user-generated content.
• Debt settlement: recipient must create an account to claim what they're owed (Venmo, PayPal).
• Conversation access: users must join the app to participate in a live discussion (WhatsApp, Slack, Discord).
• Billboarding: the product advertises itself through normal use — Hotmail's email signature, "Sent from iPhone", Calendly links, Dropbox URLs.
• UGC: users create and share content off-platform that drives others back to the product (TikTok, Quora, Reddit).
• PLA removes reliance on third-party volatility (Google algorithm updates, Facebook CPM swings).
• Referral programs are not PLA — artificial incentives attract reward-seekers who don't stay or compound.
• Best applied at idea-selection stage: if one of three startup concepts has natural PLA, lean into it.

Retention through state building

• State building: the more users invest in a product, the more costly it becomes to leave.
• Borrowed from video games — armor, skins, and progress create lock-in through accumulated value.
• Three types of non-transferable state:
  1. Reputation — eBay seller ratings, Yelp reviews, Airbnb host scores can't be ported to competitors.
  2. Audience — YouTube subscribers, Instagram followers, Twitch fans are trapped on-platform.
  3. Social graph — years of curating LinkedIn connections or Facebook friends can't be exported.
• Embedded infrastructure (Twilio, Stripe, AWS) creates state through switching cost, not just lock-in.
• "Moat" is usually misused — most companies just retain users slightly better than average; state building is the mechanism.

Writing for novelty

• Novelty = new + significant + non-obvious. Triggers a recognition response that keeps readers reading.
• Five novelty types:
  1. Counterintuitive — not how I'd have thought the world works.
  2. Counter-narrative — not what I've been told; the prevailing story is wrong.
  3. Shock and awe — true but unbelievable.
  4. Elegant articulation — a rich idea compressed into one precise sentence (Naval-style).
  5. Curiosity gap — raise a question, withhold the answer.
• Writing quality = novelty × resonance. Resonance = examples, analogies, metaphors, stories.
• Draft 1: find the novelty. Draft 2: add resonance.
• Map novelty density: have readers highlight "wow" sentences; compress the white space between them.
• Score ideas for novelty when you first encounter them — the feeling fades, the rating doesn't.

Topic selection and follow-through

• Choose topics on two axes: what you can finish, and what will be high quality.
• Pick a clear objective before writing: prove the status quo wrong, articulate what everyone thinks but no one says, contribute original research, tell a suspenseful story.
• Pair the objective with a motivation: get something off your chest, solve a nagging problem, geek out about something you want others to share.
• Objective tells you when you're done. Motivation keeps you going.
• Avoid writing about trends or news — prioritise evergreen topics worth updating years later.

The creativity faucet

• Shared process identified across Neil Gaiman, Ed Sheeran, and John Mayer — independent, prolific, consistent.
• Visualise creativity as a pipe: the first mile is wastewater. Good ideas sit behind it.
• The process: write every bad idea first — don't resist or judge them. Bad ideas are progress.
• Why it works: cycling through weak ideas trains pattern recognition; the brain learns what to avoid and gets faster at finding what works.
• Sequence: weak imitation → identify what makes it weak → iterate until original.
• Discipline, not talent, separates prolific creators: they simply allocate enough time to empty the pipe.