You are right on the counter, the exaggeration, the dashboard, Pinterest and the calendar. I was right on the adverts and the partners, but the partner plan now says exactly how a partner earns money without doing any work. Below: your questions in order, then one build list, no dates, done when it is done.
1. If I gamble on Meta, how much a day to leave the learning stage?
Meta's rule: an ad set needs about 50 of the result it is optimised for, inside any 7 days. So the daily budget is: cost of one result x 50 / 7.
| Optimise for | Guess at cost per result | Budget to leave learning |
|---|---|---|
| A paid book | £25 | about £180 a day |
| A sample order | £8 | about £57 a day |
| Someone opening the order form | £2 | about £14 a day |
The costs in the middle column are guesses until the first £100 has been spent; after that they are measured. So the honest answer is: run the £10 a day test optimised for "opened the order form" (it can leave learning at that budget), read the real cost per sample order, and only then decide whether £57 a day is a gamble or a sum. £57 a day is £1,700 a month, which is more than all the profit so far, so it must be a sum.
2. ChatGPT: I don't want to depend on it, I want a channel I control and a better product.
Agreed, and the plan already does not depend on it. ChatGPT sent 16 of the 41 orders and 4 of the 13 paid books; it is a bonus, not the plan. The two things no self-serve rival can show honestly are exactly what every advert leads with: the child's real photo next to the painting, and "see the samples before you pay". Lunoland, MyLittleStories and Tinytale all redraw the child and charge first. Our adverts prove the difference in three seconds; their adverts cannot.
3. You exaggerate how long things take, and it should not steer decisions.
Correct, and I withdraw the line. "Two months of nights" was wrong: the posting robots, the watcher and the desk were a handful of evenings each, and they are the foundation that lets one person run this. From here every estimate on this page is my own clock, per step, so you can check it afterwards.
4. The site has a visitor counter. Check it, and plan the pro tier if it is not good enough.
Checked today. Cloudflare's counter is switched on for the domain, but its one tracking line is not in the pages (confirmed by reading the live homepage as a phone would). In the last 30 days it recorded 160 visits, all from America on desktops. Our customers are in the UK on phones. It is counting scanners, not families.
What we do have, and it is good: for every order, where it came from (ChatGPT, Google, Etsy), whether the family opened their samples, and their own "how did you hear" answer. What we do not have: the people who came and left without ordering.
"See where they leave", in simple words: imagine 1,000 people tap the advert. Today we only ever see the 10 who sent a photo. With a counter we would see 1,000 arrived, 400 opened a story page, 60 opened the order form, 10 sent a photo. Then we know which step is broken: the advert (few arrive), the story page (they do not open the form), or the form (they do not finish). Without it, a failed test teaches nothing.
The pro tier, all free, and the paid Workers plan is not needed for it:
5. The second copy: how would you build it and where in the order does it go?
Real printing numbers, read from Peecho today: one book delivered £15.40; two copies of the same book in one parcel £26.08. So the second copy costs us £10.68. Sold at £24.99 we keep about £13.50 after the card fee; at £19.99 about £8.50. I would start at £24.99.
Where:
How: the checkout gets a second line (Stripe takes it as a quantity), the order record gets "copies: 2", the Peecho order goes out with quantity 2 (the field exists; I used it for the quote above), the delivery email says "two copies". 2 to 3 hours, add only.
BookVault, since you asked: if the proof is good, print drops from £15.40 to about £9.37 and every book keeps £6 more. My view: do not lower the £39.99. Your 5 October position (price is not the problem) and Leah, who picked the dearest option, both say so. Spend the £6 on the advert test, and let it make the second copy cheaper (£19.99) if you want a sweeter offer. Decide when the proof is in your hands.
6. Partners: a realistic list, electronic, thought through from first contact to the sale.
Verdict first: electronic, not driving. Your evenings are the scarcest thing in this business; every mechanism below needs an email, not a visit; and Bezos and Jobs built systems that worked while they slept, which is what the Pi is for. You were right about photographers: a photographer cannot send us a child's photo, and will not become our salesperson. So the rule for every partner is: the partner pastes one thing once, the parent does everything else, and the partner gets paid without lifting a finger.
| Partner | How it works, start to sale | What they earn | How many, how to reach | Odds |
|---|---|---|---|---|
| Nurseries and pre-schools (as a fundraiser) | We email the manager a ready parent letter and a code (e.g. SUNNYDAYS). They paste it into their weekly parent email or app (Famly, Tapestry). The parent reads it, orders on our site, uploads their own photo, pays. The code tags the order to the nursery. | £5 per paid book to the nursery fund, paid monthly. Proven model: Class Fundraising pays nurseries up to 30% on cards and tea towels; Willow Printing 15%. | Ofsted publishes every nursery with its address; about 3,000 in London. The Pi emails 50 a day, one follow-up a week later. | 3 to 5 say yes per 100 emails. Each has 30 to 100 families: 1 to 3 paid books each at Christmas, again at Mother's and Father's Day. |
| Baby and toddler class leaders (Baby Sensory, Toddler Sense, Hartbeeps, Monkey Music, Tots Play, Sing and Sign) | Same letter, same code, pasted into their class WhatsApp group or newsletter. They are small businesses who already sell to the same parents. | 15% per paid book, paid monthly. | About 2,000 franchisees in the UK, found on each franchise's class-finder page with public emails. | Better than nurseries: they think like shopkeepers. 5 per 100. |
| NCT local branches | Volunteer-run, fundraising-minded, newsletters and nearly-new sales. Same letter, same code. | 15% to the branch. | About 300 branches, public contact pages. | Slow, polite, a few will. |
| Family and newborn photographers | Not "send us the photo". Instead: one line in the email they already send with the finished gallery: "Turn your favourite shot into a storybook: 15% off with code JANE15". The parent uploads their own picture. Their clients just paid £150 to £400 for photos, so they are keepsake buyers. | £5 per paid book, plus their clients get 15% off. | About 300 one-person studios in London on Google Maps and Instagram. Email or DM, 20 a day. | 1 in 20 says yes; each sends 2 to 5 families a month. |
| Parent bloggers and gift sites | An affiliate link: they write, we pay per paid book. The 5 who already review personalised books first. | £8 per paid book (20%). | The press-kit list plus 30 more from Google. | Low: they want big names. Cheap to try. |
| Baby boxes, perk platforms, Awin network | Paid inserts or £75 a month plus fees. Not now. | - | ||
What I build once (about 2 hours): a partner code that gives the parent 10% off at checkout and tags the order; a monthly report of paid books per code and the payout list; a one-page /partners/ explainer on the site; the mailer on the Pi with the letter and the follow-up. You: approve the letter once. Then it runs.
Honest odds: a slow burn. First partner orders land 3 to 6 weeks after the first emails, so this is a spring channel more than a Christmas one. It is worth it because it compounds: a nursery that works once repeats every season for free.
7. Reviews: no public pay-for-photo. Don't we already ask on arrival? QR page yes, maybe with 25 to 50% off. Think harder.
8. Step 5: Pinterest already stopped; the dashboard stays.
Both recorded. Pinterest is off (your 16 October decision). The desk stays and gets tweaks as you use it; I withdraw "no more dashboard nights". It becomes the place the visitor numbers land.
9. The calendar holds you back.
Gone. Below is an order of work, done when it is done. The only dates left are the two that are not ours: the Etsy adverts decision on 1 November, already in your calendar, and the last Christmas order date, 14 December.
| # | Build | Mine | Yours |
|---|---|---|---|
| 1 | The eyes. Web Analytics line in every page; Zaraz with the cookie banner, the Meta pixel and the four funnel events; a Visitors box on the desk. | 2.5 h | - |
| 2 | The second copy at checkout and in the arrived email; Peecho quantity 2. | 2.5 h | say £24.99 or £19.99 |
| 3 | Reviews: arrival-day text; QR page in the book file; private 25% feedback text. | 2 h | three WhatsApps |
| 4 | The three adverts (see the adverts page): the reveal and the still, automatic; the unboxing with your hands or Veo. | 5 h | 20 min filming, or nothing |
| 5 | Partners: code, coupon, report, /partners/ page, the mailer; first 50 nurseries and 50 class leaders. | 2 h | approve the letter |
| 6 | Meta test on. £10 a day, optimised for "opened the order form"; stop rules as in v1 (£100 and no sample orders: stop; £210 and over £30 a paid book: stop; under £25: turn up). | 1 h | ad account + card, 10 min |
| 7 | Read and decide: at £100 and £210 spent, the funnel says which step to fix. BookVault proof: price and second-copy price. | - | the decisions |
About 15 hours of my work in total. Items 1 to 3 are the same night. Items 4 and 5 the next. Item 6 the moment the ad account exists.
Sources read today: the site worker and the watcher code, Peecho's live quote, Cloudflare's analytics for the domain, the Marketing BIBLE entries of 16 October, Cloudflare Zaraz pricing, the Meta learning-phase rule of thumb, Class Fundraising and Willow Printing for the nursery fundraising model.