Chapter 03
Pricing and packaging
Price on the unit the installer thinks in, which is a house, not a seat. Everything else follows from that and from one measured number: what a delivered pack actually costs us.
What matters
- Two bands ship today: Starter at £29 a month for 5 packs, Pro at £89 a month for 20.S3 That is £5.80 and £4.45 a pack.
- The "about a pound a pack" figure is wrong for the product as built. Measured on our own production database, renders alone average 96p a pack and a delivered pack costs about £1.85 today.S1 The pound is reachable, but only at volume or with fewer renders.
- Gross margin is 59% to 68% at the shipped prices, which is fine, and it is not the 95% a software person assumes. This is a business with a printing press in it.
- House accounts already exist in code and are the single most useful commercial instrument we have in the first quarter.S3
- Postage must be charged at cost, separately from the subscription, before any high-volume band ships. A postcard costs more to post than a letter, and a flat-fee volume band would leave five pence in the pound.S7
3.1 What a pack costs us, measured
These are not estimates from a price list. They are the sum of the cost_pennies column across the renders table on the production database, measured on 9 September 2026.S1
| Measure | Value | Basis |
|---|---|---|
| Renders in the database | 66 | measured across 18 packs |
| Total render spend to date | £17.30 | measured |
| Mean cost per render | 26.2p | measured, range 19p to 37p |
| Mean cost per render, most recent day | 20.9p | measured, 13 renders at the current settings |
| Mean render cost per pack | 96.1p | measured, range 62p to 143p depending on 3 or 4 styles |
Why the per-render figure moved. The first sixteen renders averaged 22.3p. Then the model was changed to the editing-precision variant at 2304 by 1536 and extra-high quality, which pushed the mean to 28.3p on the busiest day before settling around 21p.S1 Two lessons. The render cost is a dial we control, and any figure quoted without a date is meaningless within a fortnight.
Print and post
Stannp publishes UK per-item prices which include production, printing, postage and mailing, at second class, with no minimum order quantity.S7 These are the numbers that matter, because they are the only UK print-and-mail prices published at the volume we will actually run this year.
| Item | Free plan (£0/mo) | Paid plans | Volume band |
|---|---|---|---|
| A4 single-side colour letter | 88p | 82p | up to 3,999 |
| A4 two-side colour letter | 90p | 84p | up to 2,500 |
| A3 folded to A5 letter | £1.14 | £1.08 | up to 2,500 |
| A5 postcard | £1.04 | 98p | up to 2,500 |
| A6 postcard | £1.02 | 96p | up to 3,999 |
Two things in that table are counter-intuitive and both change decisions. First, an A5 postcard costs more to send than an A4 letter, 104p against 88p. The postcard is cheaper to render, not cheaper to post. Second, the paid plans save only 6p an item below 4,000, against monthly fees of £12, £48 and £315.S7 The £12 Starter plan breaks even at 200 items a month, so until we are posting 200 items a month the honest figure is the Free plan's 88p, not 82p.
Royal Mail direct, and why it is not available to us yet
At scale Royal Mail's own Advertising Mail is far cheaper, but the minimum volumes are the gating constraint and they are published: Advertising Mail Unsorted requires 1,000 Letters or 250 Large Letters per class, per day, per site, and Low Sort requires 4,000 Letters or 1,000 Large Letters. Partially Addressed Mail requires 10,000 items. Door to Door carries a £200 minimum charge per contract.S8
| Royal Mail option | Per item | Minimum | Relevance |
|---|---|---|---|
| 2nd Class Mailmark Advertising Mail, Large Letter, 100g | 89.2p at 250, 85.5p at 2,000+ | 250 Large Letters in one day | Barely cheaper than the bureau, and we do the printing. Not worth it. |
| 2nd Class Mailmark, Letter, 100g | from 48.3p | 1,000 in one day | Real money, and out of reach until a supplier or group is buying the volume. |
| Partially Addressed 2nd Class Low Sort | 36.6p | 10,000 | No name at all, so lower data-protection exposure. A future supplier-scale option. |
| Door to Door, up to 100g, new user rate | 13.46p | £200 per contract | Genuinely interesting. Wrong tool for a named house, right tool for an installer blanketing the streets around a job they have just finished. |
All Royal Mail rates exclude printing and enclosing, so the bureau's all-in figure remains the right number for the model. A revision to the price guide takes effect on 5 October 2026, which raises the second-class Letter Standard Unsorted rate from 74.0p to 80.0p and leaves the Large Letter Advertising Mail rates unchanged.S8
The delivered cost, put together honestly
| Scenario | Renders | Print and post | Total | Note |
|---|---|---|---|---|
| A pack today, posted, low volume | 96p | 88p | £1.84 | The realistic number for year one, and it uses the single-side letter price |
| A pack today, downloaded not posted | 96p | 0p | 96p | A large share of packs, in practice |
| A three-style pack, posted, over 200 a month | 63p | 82p | £1.45 | Drop the signature style and pay the £12 plan fee |
| A postcard, one render, posted | 21p | £1.04 | £1.25 | Cheaper to make, dearer to send. The saving is smaller than it looks. |
| DIY: printed at a bureau and stamped, 100 items | 96p | £1.44 | £2.40 | derived A4 sheet at 25.2p plus A5 card at 28.0p plus a 91p second-class Letter stamp, excluding envelopes, print delivery and labourS10 |
The DIY row settles a question. Self-assembly is worse than the bureau at any volume we will see, and that is before anybody's time is counted. Unless the pack contains something a mail API cannot produce, use the API.
Correct the £1 figure everywhere it appears. A pound a pack was derived for a single-render flyer at national volume and it was right for that. The Housewarmly pack carries three or four renders and posts at low volume, so it costs about £1.85. Anyone building a margin case on £1 is out by about 85%. Say £1.30 to £1.85, say why it varies, and say which lever moves it.
What the print-and-post line above does not yet establish. Stannp's 88p is for an A4 single-side colour letter. The Housewarmly pack is four pages of A4 landscape, which is a different product on that account and has not been priced. It may fold to A5 and post as a Letter at around £1.14, or it may need to go unfolded as a Large Letter, which costs materially more.S7 Until somebody sets the product and posts one, every total in this table carries that uncertainty. This is a half-day task and it is in chapter 08, week 1.
An open operational risk that belongs in this chapter because it costs money. The pack became A4 landscape on 9 September 2026. The dispatch gateway attaches the PDF and names no product, which was correct for a portrait A4 letter and is exactly the assumption that has changed.S2 A landscape pack fed to a portrait product comes back rotated or cropped, and the first anyone would know is a letterbox. Pick and test the landscape-compatible product before the first paid send.
3.2 The bands
Two exist in code. A third and a fourth are proposed here and are not built.S3
| Band | Price | Packs | Per pack | Cost if all posted | Gross margin | Status |
|---|---|---|---|---|---|---|
| Starter | £29/mo | 5 | £5.80 | £9.20 | 68% | shipped |
| Pro | £89/mo | 20 | £4.45 | £36.80 | 59% | shipped |
| Patch | £249/mo | 75 | £3.32 | £138.00 | 45% | proposed |
| Postcards | £99/mo | 75 | £1.32 | £93.75 | 5% | do not ship as drawn |
The Postcards row is the reason this chapter exists. At 75 posted cards a month, £99 of revenue carries £93.75 of print bill and leaves five pence in the pound. That is not a pricing error, it is a structural one: a card costs more to post than a letter, and any band that bundles postage into a flat fee at volume converts a software margin into a print margin. The Patch row is heading the same way at 45%.
The fix, and it is the one the trade already understands. The subscription buys the packs. Postage is charged at cost, per item, with the price stated. Every print bureau in the country sells this way and no installer will find it strange. It keeps our margin on the thing that scales (the render and the software) and passes through the thing that does not (the stamp). Do not ship Patch or Postcards until this split is built.
What each band is actually for
Starter, £29
The installer who posts a pack at the two or three good listings that come up on their patch each month. This is the band that gets tried, and the number that has to be smaller than what they already pay Checkatrade.S17
Pro, £89
A firm with a showroom and fitters, working a patch properly. Twenty a month is roughly one a working day, which is the rhythm to describe it in.
Patch, £249
The multi-showroom independents in chapter 05 with two or three branches. Sell it as coverage of a postcode district, not as a volume discount.
3.3 Why not per seat, and why not per lead
| Model | Why it is wrong here |
|---|---|
| Per seat | The buyer is a firm of four to twelve people of whom one uses software. Per-seat pricing prices the wrong axis and invites password sharing. It is also the model of the CAD incumbents at £1,000 to £2,000 a designer a year, and being compared to them is a fight we lose on price and win on nothing.S20 |
| Per lead | Attribution is unsolvable. The pack lands, the homeowner rings four months later, and nobody can prove which of the four things they saw caused it. Checkatrade turns £105.9m of lead revenue into a 2.5% operating margin doing exactly this.S17 Do not enter that business. |
| Per postcode, exclusive | Tempting, and it is what Keystone AI does at £249 a month for one trade per postcode.S28 The problem is that exclusivity given away before value is proven destroys value. Hold it back as a later upsell for the Patch band, priced separately. |
| Per pack, as a monthly allowance | Prices the unit the installer thinks in. Caps our cost exposure exactly, because the allowance is consumed at the render step, which is the expensive one.S3 A pack that fails before rendering costs the installer nothing, which is the right behaviour and worth saying out loud on the pricing page. |
3.4 The competitive price anchors
Three published prices bracket us, and all three are useful in a sales conversation.
| Anchor | Price | What the installer gets | How to use it |
|---|---|---|---|
| RenoLeadsS23 | £25/postcode/mo + £1.50/letter | New-homeowner data and a generic trade template through the door | The direct comparison. Same channel, same trigger, no picture of their room. Our per-item cost is lower and our creative is not a template. |
| MoverAlertsS9 | 84p per mailer | The only published all-in mover-mailing price in the UK market | The floor. It is what a mailer costs with nothing bespoke in it. |
| Checkatrade GrowthS17 | from £59/mo | A listing and shared leads, twelve-month fixed term | The budget we are competing for. Pro at £89 is more; Starter at £29 is half. Lead with Starter. |
| TailoredQuoteS27 | from £35/mo | AI before-and-after inside a branded quote PDF | Proof the trade pays for renders. Different moment in the funnel, so not a competitor, but it settles the "would a fitter pay for this" question. |
| Keystone AIS28 | from £249/mo | AI lead reply and follow-up, one trade per postcode | The ceiling. Somebody is already charging kitchen fitters £249 a month for answering the phone. |
3.5 House accounts
A company row carries a complimentary flag that bypasses purchasing entirely: unlimited packs, and the billing page says the packs are on us. It is toggled by one artisan command.S3 This is already built, and it is the most commercially useful thing in the codebase for the next ninety days.
| Who gets one | Why | What we ask for in return | Cap |
|---|---|---|---|
| The first ten design partners | We need packs in real letterboxes and response data more than we need £290 a month. | A named contact who will answer the phone, permission to use the firm's name, and the response numbers. | 10 firms, reviewed at week 12 |
| A supplier or manufacturer running a trial | Chapter 08. If a brand is paying for packs on behalf of its dealers, the dealers must not also be billed. | A written trial scope and a named sponsor. | One trial at a time |
| A trade body or buying group pilot | BIFIS, KBSA and the buying groups reach the exact population we cannot find a list for.S14 | An introduction to their membership and a case study. | One group |
| Anybody who asks nicely | No. A comped account given without a stated exchange is revenue given away and a customer who will never convert. Write the exchange down before the command is run. | ||
The discipline that makes house accounts work. Every comped account has an end date and a stated thing we get. Put both in the record when the flag is set. An account that is still comped at month nine with nothing written next to it is a customer we have trained not to pay.
3.6 The postcard as a volume SKU
An A5 landscape postcard, two sides, 210 by 148mm plus bleed, exists as a drawn prototype in the repository. The front is the pack's cover at half the size; the back is the postcard division everybody already knows how to read.S2 It is not a shipped product and it should become one.
Why it is the volume product
- It costs less to make. One render instead of three or four takes the compute from 96p to about 21p.S1
- It is the format the channel already runs on. Every mover-mail bureau in chapter 02 sells letters and cards, not four-page packs.S21
- It reads faster. A card is seen; a pack has to be opened.
- The format is certain. An A5 card has a published bureau price. The four-page landscape pack does not, yet.S7
It does not cost less to post. An A5 postcard is £1.04 against 88p for an A4 letter.S7 The whole saving is in the render, and it is about 75p, not the pound most people assume.
What it gives up, honestly
- One style, not three. The pack's best moment is the spread where the same room appears three ways. A card cannot do that.
- No measured plan page. The page that convinces the sceptic does not fit.
- Less room for the installer's name. The pack is a brochure with their brand on it. The card is a card.
The recommendation. Sell the pack as the hero and the postcard as the follow-up, in that order, at a stated per-item price rather than inside an allowance. Do not lead with the postcard, because the pack is what makes an installer say yes in a meeting and the card is what they use once they believe it.
Two things must be settled before the postcard ships. First, the bureau product and price for an A5 double-sided card, which is not on the account yet and is therefore an unconfirmed line in the cost table above. Second, whether a card is enough to carry the AI-visualisation disclosure and the source-photography credit legibly at that size. Chapter 11 says both must be on the artefact, and small print that cannot be read is not disclosure.
3.7 Discounts, trials and the things not to do
- No annual discount in year one. We do not yet know the churn rate, and an annual prepay locks in a price we may need to change.
- A free trial, not a free tier. Fourteen days with a hard allowance of two packs. A free tier on a product with a real marginal cost is a subsidy with no end date.
- Never discount the list price to close a deal. Move the band or add packs. A discounted price is a permanent price.
- Do not print the discount policy on anything a customer sees. It was one of the findings against the previous guide in this series and it applies here.
- Charge for postage separately once volume arrives. Said twice in this chapter deliberately.