| name | building-in-public |
| description | Grow an audience-led indie product as a designer/maker — build in public from the prototype stage, problem-first demos, the marketing↔product loop, company-of-one economics. Use when launching a side project, building an audience for it, or operating as a solo maker. Devtool landing pages/PMF → devtools (developer-tool-gtm); OSS → devtools (oss-strategy). Triggers: build in public, indie product, side project, solopreneur, launch strategy. |
Building in Public
Sources/gaps: references/sources.md — 6 Config talks + Valio Con + Show Your Work! + Dive Club; platform/audience candidates in references/coverage-gaps.md.
The task: you're a designer/maker launching your own thing — build the audience while building the product, validate before you ship, and make it pay as a team of one. (Devtool landing pages/PMF → devtools (developer-tool-gtm); OSS promotion → devtools (oss-strategy); onboarding flows → user-onboarding.)
Work the four stages below in order. Each rule is one line here; full quotes, receipts, and per-speaker texture live in references/field-notes.md, and the pre-algorithm-era bootstrap corpus (tagged Valio) in references/valio-field-notes.md.
Stage 1 — Validate before you build
- Validate at the prototype stage (Perkins): "you can start working towards product-market fit in the prototyping stage" — the loop runs on Figma prototypes and comment sections, no build required. Receipts: comment-to-launch in 32 days; ~100k waitlist emails from speculative concept posts.
- Spreadsheet before code (Garland, Valio) — model the money side first; the math surfaces the real constraints before you dive in.
- Pre-sell before you make (Long, Valio) — a pre-committed audience forces the finish.
- Sell the render before the product exists (Di Genova) — audience response to a 3D render proved demand before any tooling was paid for.
- Fake what can be faked; never open a ship gap (Garland, Valio) — his invite codes "didn't actually do anything"; the six-month finish-to-launch hiatus was his biggest mistake.
- Don't count on press (Garland, Valio) — bootstrapped products earn attention through the product, not coverage.
- The financial-viability line has fallen (Ouriach; Ridd's label: the niche economy) — production got cheap, distribution didn't; thousand-person problems are now open, so keep scope tight.
- Audience is the remaining bottleneck (Goldin): "building software isn't as hard as it used to be, but growing an audience is as hard as it always has been" — the product is no longer the moat; budget effort accordingly.
- Ship faster than feels ready (Saran): "more features don't get you closer to product-market fit. Once we started shipping faster, we learned faster."
Stage 2 — Post: rough, problem-first, findable
- Rough beats polished (Perkins; Zhang's corroborating number: a raw team reel got 150× the views of the polished one): "authentic work-in-progress vibes get so much more engagement than perfectly curated content." Post as "me," not the faceless "we."
- Problem first, then a fast visual solution (Perkins) — people move when they understand the problem and see the solution demonstrated quickly.
- Persona × cohort targeting (Perkins) — a broad persona is too shallow; intersect it with a shared-interest cohort, aim each post at one subcommunity, and trend-jack deliberately.
- The build's exhaust is the feed (Ruiz) — debugging recordings, half-working demos, and dev chaff are zero-marginal-cost posts; no separate content production line.
- Post the live edge, repeatedly (Ruiz): "you own the now and then you you own the now again" — don't hoard a polished backlog.
- Daily dispatch (Kleon) — after the day's real work, share one small process artifact; set a timer so sharing never eats the work.
- Be findable while getting good; join a scenius (Kleon) — find the community, notice what's missing, fill that void in public before worrying about monetization.
- Posting is proof-of-work (Geoco) — sharing your workflow and tools publicly is product discovery and career discovery at once; public hybrid-skill work can mint a role ("brand engineer") before the org knows it exists.
- Attribution is context; don't become human spam (Kleon) — credit like a museum label; be a fan first, an open node, a connector.
Stage 3 — Run the loop: audience ↔ product
- Optimize early content for feedback, not downloads (Perkins) — comments are the early-stage metric; "feedback brings more feedback, more eyeballs, more people to your product."
- The two-way loop is the point (Perkins): "the marketing has educated the product and the product is constantly educating the marketing."
- The embarrassment loop self-corrects the product (Ruiz) — posting pressure doubles as a quality forcing function; the loop running at the API level.
- Make everything feel like it emanated from the community (Zhang, via Yuhki Yamashita) — users are simultaneously the research channel and the marketing channel; narrowcast intimacy over broadcast polish.
- Design decisions can be marketing artifacts (Saran) — one opinionated, screenshot-able decision (the jumbo button) can out-market a campaign.
- Position to where people already are (Saran) — the one-click-login framing succeeded where password-hygiene preaching failed.
- Position by stalking your commenters (Di Genova) — let the audience that actually showed up define the category.
- Let the community set the price; defect → limited edition (Di Genova) — sealed-bid auction pricing; the "Fallout Edition" recovery converted 94% of pre-order backers.
- Tell the work's story; run the story filter (Kleon; Flarup, Valio): "our work doesn't speak for itself" — problem → work → result; and ask whether the choice would make a great story.
Stage 4 — Monetize and operate as a company of one
- Watch profit margin, not revenue (Vallaure) — solopreneur margins run "usually way above 90%"; don't benchmark against funded-startup revenue narratives.
- Time is the scarce asset, not funding (Vallaure) — design the business so "it scales with us and not away from us"; automate the decoupled parts.
- Own the core (Vallaure) — no-code is fine as long as you understand what's under the hood.
- Satellite products (Vallaure) — one shipped product surfaces adjacent demand; aim for a mini portfolio.
- Stock and flow; own your turf (Kleon) — convert daily flow into durable stock on a domain/list you control; platforms create reach, owned surfaces preserve memory and monetization.
- Teach what you know; monetize with consent (Kleon) — teaching creates demand rather than giving away the store; collect email with permission and "treat it with respect."
- Train / Trust / Trade (Di Genova) — every capability gap: learn it, hire it, or barter it; trade is the bootstrapper's hidden budget.
- Boundary mechanics are survival infrastructure (Mielke, Petty — Valio) — no late-night email; "more time does not equal better work."
- Expect the unglamorous ratio (Sheldon, Valio) — it's 98% customer-service Friday nights, not the 2% press glamour.
- Side projects compound (Smith) — the value arrives "in the opportunities that it creates"; couple products to durable principles, not a single tool.
- Cadence maxim (Perkins): "Passion will get you started, persistence will get you to the turning point, and patience will keep you going."
The counterweight voice (Aubrey Johnson, Valio)
Hold against this skill's build-for-yourself leanings: building for yourself "sometimes I think feels a little bit lazy"; investors weigh traction > team > product; and the purist consolation is fake — "if the product fails and you didn't raise money you still failed it's still zero."
Checklist
Relationship to other skills
devtools (developer-tool-gtm) — landing-page anatomy, conversion mechanics, and PMF measurement when the product is a devtool (much of its copy/conversion guidance generalizes; this skill owns the audience/content/solo-economics side).devtools (oss-strategy) — the sibling playbook when the artifact is open-source code (READMEs, promotion loops, maintainer economics).user-onboarding — owns the funnel itself (sell the dream, quick wins, share artifacts, ask-after-value); this skill fills the funnel.hardware-product-design — when the indie product is physical: suppliers, tolerances, inventory, unit economics (Work Louder's ops side).design-prototyping — the prototype-stage validation mechanics (what to fake, how to demo); here they double as content.shape-up — appetite/betting once the indie product grows a team.
Staleness note (Kevin's rule): auto-captioned 2024–25 talks plus a 2014 book; platform/algorithm specifics (TikTok duets, trend-jacking, newsletter vendors, social-network examples) decay fastest — the feedback-over-downloads, persona×cohort, owned-surface, and margin-over-revenue rules are the durable layer.
