Coloring books: Shopify launch plan
Goal: sell a religious Christmas coloring book before Christmas, mainly for the learning and for her enjoyment. Profit is a bonus. Channel (founder, 2026-10-06): Shopify first. KDP is parked. The inside pages mostly carry over to KDP later; the cover would need a rebuild, and KDP requires an AI-content disclosure. Rule: no inventory. Lulu prints each copy on demand after a customer pays.
Decisions only you two can make (this week)
Status 2026-10-07 (founder iMessage 06:51-06:57 ET): 1 = DECIDED, store runs under Ray Data LLC (trade name/DBA; LLC EIN + business bank for Shopify Payments; ask Collective Hub about the S-corp return before Payments setup). 2 = OPEN, 5 options sent (Larkline Press, Hearthpages, Hollyhock Pages, Willowline Press, Linen Leaf Press); trademark check on their pick before buying the domain. 3 = DECIDED, Basic. 4 = CLEAR, paid Quiver plan confirmed; Quiver ToS 2026-03-25 allows commercial use on paid tiers, no exclusivity, we indemnify; add an "illustrated with AI, composed and reviewed by hand" listing line; check the "Improve QuiverAI" toggle.
- Whose business is it? That person's name goes on the store and on Shopify Payments (identity, bank, taxes). If it's her project, it should be in her name.
- Store and brand name. Ray will draft 5 options once the owner is decided.
- Billing: Basic. New stores currently get 3 days free, then $1/mo for 3 months, which covers us through Christmas (confirm at signup). In January, decide whether to cancel, pay monthly ($39), or pay annually ($29/mo).
- Rights check: confirm QuiverAI's terms allow selling the art. Purely AI-generated images can't be copyrighted in the US, so others could copy them. That's a known, accepted risk for a learning project. If any page quotes Scripture, use the KJV (public domain) and have someone you trust check the theology.
Setup (about an hour of your time; Ray does the rest)
- Create the Shopify store and choose Basic.
- Set up Shopify Payments.
- Create an app in Shopify's Dev Dashboard and install it on the store. Ray will send click-by-click steps. The credentials go into 1Password.
- Create a Lulu account and connect the Lulu Direct app (one OAuth click).
- Turn on Shopify Tax and register for a sales-tax permit in your home state.
- Publish the refund, privacy, terms and shipping policies. PDFs are non-refundable; Lulu reprints misprinted books. Ray drafts all four.
Ray then builds the theme, the product pages and the structured product data (age range, theme, page count, one-sided pages). Ray also sets up the free digital-downloads app and shipping profiles. Each PDF carries a license: personal and household use only, no resale or sharing. Every product is drafted first and published only after your OK.
First products
| Product | Format | Price | You keep (est.) |
|---|---|---|---|
| Religious Christmas coloring book, 25 designs, one-sided | Lulu paperback, 60 lb paper | $16.99 + shipping | ~$11 |
| Same book as a PDF | Instant download | $7.99 | ~$7.45 |
| Nativity packet (8 pages) | $4.99 | ~$4.55 | |
| Angels and shepherds packet (8 pages) | $4.99 | ~$4.55 | |
| Bundle: print book + both packets | Mixed | $24.99 + shipping | ~$19 |
Notes on the estimates:
- Card fees are 2.9% + 30¢.
- Lulu's $4.28 is its starting price for a Letter-size coloring book on 60 lb paper. A 52-56 page interior will cost a bit more, Lulu adds a $0.75 per-order fee, and Lulu may charge sales tax on the print cost unless a resale certificate is on file. Confirm all three in Lulu's calculator before setting prices.
- The packets are worth selling only if they hold new art that isn't in the book. Otherwise, drop them and the bundle.
- Before launch, place one real test order to confirm that Lulu fulfills the print book and the download app delivers the PDFs.
Timeline (to land before Christmas)
By Oct 12: decisions 1-4 made, store set up, credentials handed over.
By Oct 19: the 25-page book is built from the asset library, with these composition fixes:
- objects snap to the ground
- lines thin with scale but never drop below a printable floor
- no area to color smaller than about 1/4"
- figures face the focal point
Then the cover, then her page-by-page review. If 25 pages aren't ready by Oct 19, cut to 20 rather than delay the proof.
By Oct 21: order a Lulu proof. Allow 1-2 weeks.
When the proof arrives: test it with real markers and crayons and check for bleed-through.
About Nov 1-5: launch to her network first (church, preschool, friends). Ask the first buyers for a photo and a review.
By Nov 7 (milestone, added 2026-10-07 on founder's go): agent shopping audit v1. Ray builds a ShopGym-style test kit once real listings exist (estimate ~1 day, a few dollars of API per run): 20-30 shopper asks written from our products and policies, run by an AI shopping agent in a real browser against our store and 2-3 competitor stores, stopping before checkout. Scored on: found the right product, reached the cart, answered the question correctly. Fix what it fails before ads start; rerun after every theme or listing change. Same week: SimGym merchant preview for layout picks (section below). Reusable for any future Shopify store. Basis: [[2026-10-07-shopify-simgym-shopgym-simpersona]].
Nov 8-29:
- the ad test (below)
- Pinterest pins of sample pages
- check whether Muse finds the store when asked a shopper's question (plus an audit rerun after any listing change)
Last order date for Christmas: about Dec 8 is a guess. Set the real date in late November from Lulu's delivery estimates (production slows near year-end), then post it on the store.
Dec 15: scorecard.
Before ads: SimGym ($0, added 2026-10-07 on founder's ShopifyEng link)
- Shopify SimGym (merchant preview, no waitlist for eligible stores) runs AI shoppers through two versions of the storefront in a real browser and reports which converts better, in under an hour. Use it to pick theme, product-page layout and cart flow before any ad dollar.
- Caveats: shoppers are calibrated from the store's own visitor logs, so a brand-new store gets generic shoppers until real traffic arrives; it cannot test prices, discounts, checkout or campaigns, and does not replace live A/B tests.
- Eligibility for our plan tier: confirm in the admin once the store exists.
- Sources: ICLR 2026 expo talk, arXiv 2602.01443, Modern Retail, movingprimates limits writeup.
Agent channel facts (checked 2026-10-07)
- Shopify Agentic Storefronts is on by default for eligible stores (ChatGPT, Google AI Mode/Gemini, Copilot, Meta). Eligibility includes "a track record of genuine sales to real customers", so a new store may not appear until it has sales. Digital-PDF eligibility unconfirmed.
- [[2026-09-19-muse-connectors-agent-distribution|Muse]]: Shopify ANNOUNCED Shop Pay checkout for Muse on 2026-09-21; Shopify's Meta help page still says "Muse when available". Not confirmed live (corrects Ray's 10-06 "turned on across all stores").
- ChatGPT dropped in-chat Instant Checkout 2026-03-04; buyers go to the merchant checkout.
- Launch-day moves: confirm agentic channel on; refund + PDF-license pages; listings written for specific asks; ISBN on printed books; first real sales + reviews; monthly prompt checks (the Nov 7 audit).
Ad test (the demand monkey)
- Facebook/Instagram ads promote the print book and bundle only. A $4.99 PDF can't pay for ads.
- Meta has not allowed religion-based targeting since 2022, so target coloring, Christmas crafts and homeschool interests instead. Ad copy must not assume the viewer's faith.
- Cap: $75 total.
- Kill line: after $40 spent, stop the ads if the cost per purchase is above $10. Organic channels continue.
Budget
Shopify plan: ~$3 during the promo, outside the $200 envelope. After that, $39/mo only if we keep the store (paid by cancelling ConvertKit and/or the car wash subscription, founder 2026-10-06).
Learning envelope, capped at about $200:
Item Amount Art ~$15 Lulu proofs + test order ~$30-40 Ads up to $75 Buffer the rest Spent so far: ~$4.38 (bake-off + composition test).
Scorecard (Dec 15)
- Her: did she enjoy it, and is she proud of the book?
- Market: purchases from people outside her circle, and ad cost per purchase.
- Agents: did Muse surface or sell it at all? Even a "no" is a [[DESIGN-scribble-works|Scribble Works]] lesson.
What Ray never does
Move money, issue refunds, change payouts, or change a price or publish a product without your OK.