Marketing without a budget

Thirteenth post in “Tools I build with.” Everything in the previous four posts (the Printful store, the artistic direction, wearlunaire.com, the wiring) was the easy part. The hard part is whether anyone shows up. A solo-built brand with a 999-piece edition has roughly the same merchandising stack as a 90s indie label and roughly the same problem as a 2026 Substack writer: the product exists, the audience does not. This post is honest about marketing budgets, dishonest comparisons to avoid, and the zero-budget tactics that actually move a needle when you do not have one.

The Bieber line of comparison everyone should stop making

Justin Bieber’s clothing label SKYLRK generated $15 million in merchandise sales across the two weekends of Coachella 2026. The previous all-time festival merch record was around $1.7 million. SKYLRK’s weekend-one alone (around $5 million) tripled it. The brand operated a dedicated SKYLRK Shop, a branded SKYLRK Oasis cooling space for festival-goers, and a parallel online drop. Bieber was paid $10 million for the booking itself, the highest in Coachella history.

That is the comparison every founder of a small label has read this month, and it is the comparison every founder of a small label should ignore. The numbers are real and they are uninstructive. SKYLRK did not launch in days; it arrived in front of a 200,000-person audience that had already paid to be there, on the back of two decades of pop infrastructure, with $10 million of activation budget paid by someone else. A solo-built numbered crewneck on the open internet is not the same problem.

The useful read across is the opposite of the obvious one. SKYLRK had three things working: a captive audience, a physical retail moment, and a scarcity story (festival-only availability for weekend one). The captive audience is the one nobody starting from scratch has. The other two (a moment, a scarcity story) are within reach of any small brand willing to be deliberate about them.

The actual budget reality

A realistic marketing budget for a solo-built first edition, in 2026, sits between €0 and €500. Some founders have more, but spending more on a first launch is usually a mistake (the spend goes against an audience that does not yet know whether to trust the brand, which is the worst phase to buy attention). The whole budget question deserves an honest framing.

The headline is that marketing spend has decreasing returns at the bottom of the curve. Spending zero is fine; spending one to five thousand poorly is worse than spending zero. The point of this post is the €0 column.

Print-on-demand is the safe route, marketing is what makes or breaks it

The whole point of episodes 09 to 12 is that the production stack is a solved problem for a solo builder. PoD plus a custom site plus an API token is a configuration job, not an art project. The product exists in days because the tooling exists.

What does not exist in days is an audience. Marketing is the only part of this whole journey that does not respond to better tooling. There is no Printful for distribution. There is no API token for trust. The whole thing stands and falls on whether the brand finds the right thirty to fifty people first, who then find the next thirty to fifty.

The honest read is that a small brand using PoD for production is making the safe choice, the cheap choice, and the slow choice on the production side, because all three of those let the slow part (marketing) breathe. Inventory pressure forces a launch on someone else’s timeline. PoD lets the launch happen when the audience is ready, not when the warehouse needs clearing.

The waitlist as marketing primitive

A waitlist on a small brand is usually treated as a data-collection form. The interesting move is to treat it as a marketing primitive. The mechanics of the Lunaire waitlist (claim a number, three digits, only 999 ever) does three things at once if you let them.

• Low numbers become bragging rights. The first ten reservations carry status that the next thousand do not. The early adopter who claims #003 has a reason to post about it. The brand has nothing to do with that posting; the system is set up so the buyer wants to.
• The reservation itself is shareable. A confirmation page that shows the buyer’s number large in the brand wordmark is a screenshot worth taking. A receipt with a serial is content. The waitlist UX is the shareable artifact.
• The cap is a countdown. When the waitlist is at 200, the brand can say so. 200 of 999 claimed is honest scarcity. Almost gone, last few left is dishonest scarcity. The numbered system makes the first option available without ever needing the second.

None of this requires paid acquisition. The system carries the marketing load if it is built to.

Pick one platform, go deep

The instinct on launch day is to be everywhere at once. The honest tactic is to be on one platform deeply, and to copy-paste the rest as background.

For a numbered piece of soft luxury sold at €79.99, the audience map looks roughly like this:

• Instagram, primary. Still the right home for clothing in 2026, especially clothing that photographs well in low-light tonal contexts. The buyer reads here, the references read here, the press reads here. One feed account, occasional Reels, no Stories at first.
• TikTok, secondary. Behind-the-scenes content (the build, the embroidery test, the draft order in the dashboard) does well; product-flat-lay does not. Different audience, different format, run only if you can produce a video a week without hating it.
• LinkedIn, off the list. Lunaire is a luxury garment, not a B2B tool. LinkedIn does well for the journal in this site (a hiring portfolio audience), but does not work for selling sweatshirts. Greying out the LinkedIn icon in the launch plan is fine.
• X / Threads, optional. Useful for the tool-builder audience and the tech press, useless for fashion buyers. Cross-post a build update once a week and ignore it the rest of the time.
• Substack / newsletters, the dark horse. A small newsletter mention from someone whose taste matches the brand outperforms most paid placements. A direct outreach DM with the launch story attached is €0 and might land.

One platform deeply, two more on minimum viable, the rest off. The discipline matters more than the choices.

Content cadence when there is nothing to ship yet

Pre-launch is the worst phase of the whole journey for content. The product is not in hand, the photography is mockups, the story is “please trust us.” The fix is to stop trying to be a brand that already exists and lean into the build.

• The journal is content. This post and the four before it are the marketing too. They explain the product, the price, and the thinking behind it. A buyer who has read three of them is closer to #042 than a buyer who has only seen a flat-lay.
• Process beats product. The Printful Design Maker screenshot, the rendered #888 PNG, the draft order in the dashboard, the sleeve mockup with tonal contrast. None of these are products. All of them are content. People who want to follow the build will share these in a way they will not share another flat-lay.
• One real piece of work per week. Not seven posts. One real one. A 60-second video showing the embroidery file rendered, then stitched, then on a body. The other six days are reposts, comments, replies. Volume is the trap; consistency at low volume is the move.

Fun zero-budget tactics that actually move

The lazy version of zero-budget marketing is “post on socials and hope.” The interesting version is closer to the SKYLRK move scaled down: pick one moment, one mechanic, one community, and do something specific.

• Soft-launch the first 50 numbers to friends. The first fifty reservations should go to people who will actually wear the sweatshirt and post about it. This is not a PR favour, it is a quality-control gate: the wrong fifty wearers send the brand the wrong direction. The right fifty wearers do the early marketing for free.
• Trade your skills for distribution. Anyone reading this journal can build a small free tool for a community they want access to. A Notion template for an astrology newsletter, a Figma kit for a fashion-design Discord, a small generator for an art collective. The trade is: free tool, in return for a one-line mention in their newsletter or pinned post. This is genuine swap, not influencer outreach. It works for the same reason newsletter sponsorships work, at none of the cost.
• Drop on a slow news weekend. Fashion month, awards season, and major holidays are crowded. A first launch fights for attention against ten thousand bigger ones. A boring Tuesday in late June is open ground.
• Make the build itself the moment. SKYLRK had Coachella; a solo builder has the launch livestream. A two-hour Twitch stream of the founder placing the first ten draft orders, talking through the design, showing the Printful dashboard. Fifty viewers is a real audience for a 999-piece edition. The recording becomes ten clips for the next two weeks.
• Write the cold email you would actually send. Five small newsletters whose taste matches the brand, five micro-press editors who cover small labels, five micro-influencers whose feed already looks like the brand. Each gets a personal email with the journal link, a sample image, and a number reserved for them if they want it. Most ignore the email. One responds. That one is enough for a first launch.
• The sample-to-influencer trade, but small. A free sweatshirt at wholesale cost (€43) sent to the right fifty-thousand-follower micro-influencer outperforms a €5,000 paid post from the wrong million-follower one. Five samples is €215. Pick the five carefully.

What to be aware of

• Discount loudly is hard to reverse. A first-launch 20% off launch code looks like marketing and acts like a margin cut. The brand is now training every future buyer to wait for the next sale. Save the discount lever for a v2 you do not yet need.
• Honest scarcity beats fake scarcity. 200 of 999 claimed is true and motivating. Last 3 left, sale ends in 47 minutes is the dishonest version, and it ages badly. The numbered model gives you the honest one for free.
• Conversion lags reach by weeks. Someone who first encounters the brand today does not buy today. They lurk, they read another post, they see a friend wear it, and they buy in week six. Counting day-one sales as the verdict on day-one marketing is the most common mistake. Patience is not a personality, it is a planning input.
• Founder presence is the cheapest distribution. A founder who replies to comments, ships handwritten thank-yous, and answers DMs personally outperforms most paid stacks at this scale. It does not scale to ten thousand orders; it does scale to the first ninety-nine, which is what actually needs to happen first.
• The hardest sale is the first ten. After ten real wearers exist, the next ninety are easier. After a hundred, the next nine hundred are easier still. The marketing problem is asymmetric: the first ten are the whole problem, and they cost time, not money.

What’s next

• Wire payment. Stripe Checkout (or Mollie) on top of the existing waitlist, the trigger that flips a Printful draft to a confirmed order, and the receipts engine for the first real wearers.
• Photograph the first piece off the line. Tonal embroidery on a real body in real daylight, not Printful mockups. The brand starts the day a real one ships.
• Open the journal up. The next entries cover what actually happens once buyers exist: returns that should not happen (personalised goods are exempt from EU 14-day, episode 05b), the surprises in real fulfilment, and the v2 questions a working brand has to answer.

Subscribe below to follow the rest of the launch.