Grow by writing what you want to read — not what you think the audience wants

Claim

The most reliable strategy for building a content audience is to write what you want to read — not what you think the audience wants. Self-oriented standards filter for content that has genuine emotional engagement and authenticity, which readers detect as real rather than manufactured. Writing for an imagined audience produces hedged, optimised, voiceless content that nobody — including the writer — would seek out.

Mechanism

"Audience-led" content discipline sounds correct but produces a predictable failure mode: the writer optimises for what they imagine the audience wants, which is usually the lowest-common-denominator version of what they themselves would value. The optimisation removes the writer's specific taste, distinctive perspective, and unusual angles — the very properties that make content worth reading. Self-oriented standards reverse the failure: if the writer wouldn't read it, neither would anyone else. Handley grew Total Annarchy from zero to 42,000+ subscribers using exactly this discipline — defining a narrow niche, sticking to her own taste, and removing the optimisation pressure that would have diluted the voice. The principle generalises: the fastest-growing newsletters, podcasts, and bodies of work are typically the ones whose creator was the first reader.

Conditions

Holds when:

• The creator has developed a clear personal taste and a defined niche.
• The creator's taste aligns with at least one viable audience segment (you can find them).
• The creator has patience — growth via this approach is often slower than audience-optimised approaches but more durable.

Fails when:

• The creator's taste is unique to themselves and aligns with no real audience — narcissism dressed as creative discipline.
• The creator lacks self-awareness about what they actually enjoy vs. what they think they should enjoy.
• The category requires audience-led optimisation by structure (some commerce categories, some B2B content with specific customer journey requirements).

Evidence

"writing what she wanted to read"

— see raw/expert-content/experts/ann-handley.md line 17.

Signals

• The creator can name what they find interesting in their own content vs. what they imagine the audience wants.
• Engagement metrics correlate with the creator's stated favourite pieces, not with the pieces they expected to perform best.
• Audience growth is steady and durable rather than spiky or paid-acquisition-dependent.

Counter-evidence

"Write what you want to read" can curdle into self-indulgence — content that serves only the creator and finds no audience. The discipline requires the creator's taste to overlap with at least one real audience segment, and to refine the overlap over time. For very-early-stage creators with no track record, audience research can usefully complement self-direction.

Cross-references

• In an AI-flooded content market, voice is the only defensible advantage — distinct, authentic, sounds like one source — voice is what self-oriented writing produces; this card is the discipline that develops voice.
• The "So what?" step is the most-skipped move in content creation across B2B and B2C — Handley's foundational claim; "so what?" is the test that catches self-indulgence.
• The goal isn't to maximize numbers — it's to be missed if you stopped. Find the smallest viable audience. — Godin's adjacent claim; the smallest viable audience is the segment that matches the creator's taste.