Little Tailor Books: how to get more orders

Written 15 October 2026 from the sheet, the desk, the Marketing BIBLE and what was checked live today. Every number here has a source; nothing is estimated unless it says so.

The short answer

The shop is growing on its own. Every advert we bought lost money because the sums never worked, not because of bad luck. The one channel every big name in our market lives on, we have never tried.

Growing: 2 to 3 orders a week in August, then 7, 8 and 6 in the last three full weeks. About 1 in 4 families who get the free samples go on to pay. Each paid book leaves about £19 after printing, card fee and painting.
Why adverts lost: Google and Etsy adverts are "search" adverts. A click cost 70p to £1.10 and we needed 25 to 50 clicks for one paid book. That is £20 to £50 of adverts for a book that keeps £19.
Never tried: Facebook and Instagram adverts (Meta). Wonderbly has 51 adverts running in the UK today, Librio 23, Hooray Heroes started a new set on 9 October. Our "one photo becomes a painting" video is exactly the kind of picture that stops a parent scrolling. Also never tried: partners who are paid only when a book sells.
Honest odds: about 1 in 3 that Meta adverts pay for themselves at the first try. We will know by mid November, for about £250. If it works it is the first thing we have ever had that we can simply turn up.

Where we are, in numbers

Orders (free samples) since August

Paid books

Kept per paid book (last 10)

£18.78

Samples that turn into a sale

1 in 4

3 Aug

7 Sep

5 Oct

sample orders paid, per week (last three full weeks: 7, 8, 6 orders)

The big costs so far (£1,086 out since July) were foundation: painting styles, demo books, repaints, the robots. They were built once. The running cost of a book is £15.57 print + about £3 painting + about 80p card fee.

1. What we have tried, and what it brought

Channel	Cost	Result
ChatGPT and AI answers	£0	16 of the 22 known arrivals, 4 paid	best
Word of mouth, no tag	£0	21 arrivals, 9 paid	good
Etsy shop (no advert)	£0	2 orders, both paid, 0 book reviews	tiny
Google search	£0	24 clicks in 28 days, 4 arrivals, 0 paid	slow
Google Ads (August)	£150	48 clicks, 2 sample requests, 0 sales	dead
Etsy Ads (Jul to Oct)	£38	70 clicks, 1 sale. £38 of advert per sale	loses
Facebook page + Instagram (daily reels, robot)	£0	0 and 1 followers, about 23 views a reel, 0 orders	empty room
Pinterest (200 pins, Oct)	£0	very few views, 0 orders	too early
Gift guide pitches (22 + 18 follow-ups)	£0	0 features	closed
Meta adverts (Facebook + Instagram)	never	never tried	the gap
Partners paid per sale	never	never tried	the gap

Sources: the desk facts rebuilt today 14:37, the Marketing BIBLE entries of 20 Aug, 1 Sept, 7 Oct and 15 Oct, Search Console 15 Sept to 13 Oct, the Meta Ad Library read today.

2. Why the adverts lost money (the sums)

One paid book needs about 4 sample orders. Nobody knows how many clicks one sample order needs, because the site has no visitor counter. A fair guess from Google's numbers is 10 to 20 clicks per sample order. So one paid book costs 40 to 80 clicks.

Search adverts: 70p to £1.10 a click. 40 to 80 clicks = £28 to £88 per paid book. We keep £19. It can never work.
Meta adverts are priced differently: about £6 to £15 to show the advert to 1,000 people, and a click is often 40p to 80p. With a picture that stops people, 1 in 50 to 1 in 100 who see it click. That is the only reason the sum can come out under £25, and it is not certain.

Three things said straight, because you asked for it:

Two months of nights went into robots posting to an empty room, dashboards and a 0-for-22 press pitch. Zero nights went into the channel the whole market uses. That is on me.
The site has no visitor counter, so we cannot see where people drop off. The plan fixes that this week.
14 review asks, 0 reviews means the ask is broken, not the families. The plan changes the ask, not the number of asks.

3. The plan: five steps

Each step says what, who, cost, when, and the rule for stopping. Steps 1 to 3 are the channels. Step 4 is reviews. Step 5 is what we stop.

Step 1. Make each customer worth more this week, £0

Why first: adverts only work if a customer is worth more than the advert. Today a customer is worth £19. These two changes lift it to about £27 without touching the price.

"A second copy for Grandma or Grandad, £24.99." Offered on the approve screen and in the delivered email. Printing costs £15.57, nothing new to paint, so about £8 extra on every family that says yes. Grandparents are the Christmas buyers. I build it, add only, nothing existing changes.
"Another story for {child}" email to the 13 paying families and the 30 sample families: their photos are already on file, so a Christmas book needs no new upload. One email, sent by the Pi, you approve the words once.
Price stays £39.99 to Christmas. Changing price mid-season muddles every test. £44.99 is a January test.

Step 2. Meta adverts test on by 1 Nov, cap £250

What: Facebook and Instagram adverts to UK parents and grandparents, showing the 15-second "photo becomes a painting" reel and the four Christmas books. The click lands on the Christmas page.

I build: the Meta pixel on the site (counts clicks, sample orders and payments; also the visitor counter we never had), a small cookie yes/no banner the law needs before the pixel can run, three adverts from reels and pictures we already own, the campaign itself.
You do (10 minutes): open the ad account under the existing Facebook business and put a card on it. I cannot do that part.
Budget: £10 a day for 21 days = £210, plus £40 reserve. Not a penny more without a new decision.
Stop rules: £100 spent and 0 sample orders: stop. £210 spent and a paid book costs over £30: stop. Under £25: turn it up to £20 a day until 14 December.
The rule I am changing, openly: the Marketing BIBLE says no channel may cost more than £15 a customer. For this test the line is £25, because step 1 makes a customer worth about £27, and because at £10 a day the campaign never leaves Meta's "learning" stage, so the test understates what a bigger budget would do.
Odds: 1 in 3 first time. Most likely failure: people click, look, and do not order samples. Then we at least see where they leave, which today we cannot.

Step 3. Partners paid only when a book sells November, £0 upfront

What: anyone who sends us a buyer gets paid per paid book, nothing for clicks. Two kinds:

Parent bloggers and gift sites: £8 per paid book (20%). The 22 we pitched for free got nothing from us; with money in it some will list us. This is how most gift guides actually work.
School and nursery photographers: they already hold the photo and the parents' trust. A code at 15%, they share it with every family on photo day.
I build: the partner link and code (the "came from" tracking already exists), a monthly list of who is owed what, and the outreach emails sent from the Pi, 20 a day. You approve the template once.
Odds: slow and real. 1 to 3 active partners by Christmas, each 2 to 10 books a month in season. It scales without your time.

Step 4. Reviews this week, £0

Your idea, a refund for a 5-star review, the facts, then your call as you asked:

In UK law since 6 April 2025 an incentivised review that hides the incentive is a banned practice (Digital Markets, Competition and Consumers Act). The CMA fines directly, up to £300,000 or 10% of turnover, whichever is higher.
Etsy's own rules call "compensation for a positive review" extortion and can suspend the shop. Trustpilot names refunds as a banned incentive and removes the reviews.
The two places where stars actually bring new buyers are Etsy search and ChatGPT's product list, which takes its stars from Etsy. Those are exactly the two that would delete us for it. So my advice is no, and not for the ethics.

The legal version that gets most of the effect:

Pay for the photo, not the stars. "£10 off your next book for a photo of {child} with the book and one sentence." It goes on our site labelled honestly ("families who sent a photo got £10 off"), and it feeds the adverts in step 2. Allowed, because the incentive is disclosed and it is not a star rating.
Ask on the day the parcel arrives, by text. The Pi already knows the arrival day from Ship24. A text with one tap to Trustpilot (website buyers) or Etsy (Etsy buyers) beats an email the next morning.
A printed page at the back of every book: "Did {child} love it? Scan to tell us." A QR code, added to the book file, so the ask lives inside the thing they are holding.
Expect about 1 in 8 asks to become a review. At today's pace that is about one review a week, and the first five matter most.

Step 5. What we stop spending time on

Facebook page and Instagram reels: the robot stays on, no more fix-time.
Etsy Ads: the 1 November rule stands (off unless an advert order appears).
Gift guide pitches: closed.
Pinterest: I build batch 2, you upload it once (10 minutes). Nothing more until it shows an order.
No more dashboard nights. Nights go to steps 1 to 4.

4. The calendar to Christmas

Last Christmas order date on the site: Sunday 14 December.

When	What happens
16 to 25 Oct	Second-copy offer live. Pixel + cookie banner + visitor counter live. Three adverts built. Partner link built. Arrival-day review text + photo offer live. Pinterest batch 2 file to you.
26 Oct to 1 Nov	You open the ad account. Meta adverts ON at £10 a day. Partner emails start, 20 a day. Etsy Ads decision 1 Nov as already planned.
8 Nov	First read: £100 spent. Clicks, sample orders, where people leave. Stop if 0 sample orders.
15 Nov	Second read: £210 spent. Cost per paid book. Stop, hold, or turn up to £20 a day.
16 Nov to 14 Dec	Run what works. Partners' first orders land here if they land.
January	Price test £44.99. TikTok Shop with creators paid per sale (bigger build, not before Christmas).

5. What I need from you

Yes or no to the plan, and to the £25 line for the Meta test.
Reviews: the legal version, or not.
When I say the adverts are ready: 10 minutes to open the ad account and add a card.
Pinterest batch 2: upload the file when it arrives.

Everything else in this plan is built and run by me on the Pi, add only, nothing existing changed.

6. Honest odds, all together

Doing nothing: 6 to 8 orders a week, about 2 paid, so roughly 15 to 25 paid books over November and December.
With the plan, likely: the second-copy offer and the repeat email add a handful of books for free; reviews start arriving; Meta gives us a real number for the first time. Christmas lands around 25 to 35 paid books.
With the plan, if Meta works (1 in 3): 10 to 30 extra paid books in November and December at £25 or less each, and a dial we can turn up in January. Christmas 40 to 60 paid books.
Worst case: we lose £250 and learn where people leave the site, which we have never been able to see.

Sources read today: CMS Law on the DMCC Act and fake reviews, Etsy House Rules, Trustpilot guidelines, Meta Ad Library, Wonderbly UK, UK Meta ads cost benchmarks.