---
name: convention-retail
type: convention
license: CC BY-NC-SA 4.0
description: >-
  The shared retail and e-commerce model for Aled's ventures — the customer
  touchpoint chain from discovery to repeat and referral and what each step must
  do, the retail KPI vocabulary beyond convention-shopify (gross and contribution
  margin, CAC, LTV, LTV:CAC, sell-through, return rate, repeat rate, email revenue
  share, AOV levers) with where each is measured, the channel model and when each
  channel earns its place at 0→1 versus scale, and the fulfilment models in play
  (own stock, print-on-demand via Printify, maker-made small batch) with what each
  means for margin, lead time and returns. Read before any storefront, lifecycle,
  newsletter, launch, retail-strategy or unit-economics work, and before
  playbook-store-launch, so Pixel, Reach, Sonar and the operators share one model
  rather than re-deriving it per run. Owned by Sonar; sits beside
  convention-shopify (tooling) and convention-brand (identity) without
  duplicating either. Surfaced by APP-1080; first draft 2026-09-03.
---

# Retail and e-commerce model

What good retail looks like across the customer's whole path, in one place, so Pixel, Reach, Sonar and the operators work from one model rather than re-deriving it per run. `convention-shopify` holds the tooling and KPI queries and `convention-brand` the identity; nothing held the retail judgement between them. Owned by Sonar, because the touchpoint chain is a journey map in the sense of `convention-user-modelling` and the model is discovery output first. A first draft; grown as BARA and Keroma trade.

## Where it sits

Three conventions, three jobs, no overlap:

* **`convention-shopify`** — the tool: store in scope, the connector-vs-theme split, naming, the ShopifyQL query behind each KPI, the human-action boundary. Pointed at here, never restated.
* **`convention-brand`** — the identity: positioning, voice, palette, type, imagery per venture. This convention says what a product page must *do*; the brand instance says how it looks and sounds.
* **`convention-retail`** — the model: what each touchpoint is for, what the numbers mean, which channels and fulfilment models suit which stage.

## The customer touchpoint chain

Eight steps in the order the customer meets them, each with one job; a step not doing its job is where `skill-conversion-diagnostic` looks first. The chain is the stage spine for any store's journey map or service blueprint — reuse it rather than inventing one per venture.

1. **Discovery** — earn the visit. Search, social, referral, marketplace, press, a market stall. Make the proposition legible at a glance and land the visitor on the page that matches what drew them, not the homepage by default.
2. **Collection / PLP** — help them choose. Show the range as a coherent set, surface price and variant without a click, give a reason to open one product over another. Merchandising order is a decision, not the default sort.
3. **Product page** — remove doubt. Answer what stops a purchase: what it is, what it is made of, how big, who made it and how, when it ships, what happens if it is wrong. Imagery in use, not only on white.
4. **Cart and checkout** — do not lose them. Shipping cost and delivery window before checkout, the payment methods the audience uses, the free-shipping threshold visible where one exists. Every extra field is a leak.
5. **Shipping and order comms** — keep the promise. Confirmation, dispatch, tracking, a delay notice if needed, in the brand voice. The customer never has to ask where the order is.
6. **Unboxing** — the first physical brand moment. Packaging, insert, the product presented as intended. A maker brand's strongest owned touchpoint; print-on-demand's least controlled.
7. **Post-purchase** — close the loop. A review request timed to arrival, care instructions, a reason to return that is not a discount. Returns handled without argument, on the policy stated at step 3.
8. **Repeat and referral** — the economics. Everything above exists so this happens; repeat rate and referral measure whether the chain worked.

## The KPI vocabulary

`convention-shopify` defines the headline metrics and their queries (revenue, orders and AOV, sessions and conversion, repeat-customer rate, top products, channel mix). The vocabulary below sits above those; where a query exists it is named, not restated.

| KPI | Plain English | Where it is measured |
| -- | -- | -- |
| **Gross margin** | Revenue minus the cost of goods sold (product, inbound freight, packaging), as a share of revenue. | Per product in the operator's cost sheet (or Shopify cost-per-item where entered), against the *Revenue* query. |
| **Contribution margin** | Gross margin minus the variable cost of getting each order out — payment fees, shipping subsidy, marketplace fees, attributed paid acquisition. What an order actually contributes to fixed costs. | Per order, cost sheet against the *Orders & AOV* query. The number a costed budget proposal is built on. |
| **CAC** | Acquisition spend in a period divided by new customers won in it. | Ad-platform spend against `customers` minus `returning_customers` from the *Repeat-customer rate* query. |
| **LTV** | What a customer is expected to spend over their life with the brand, taken at gross margin. At 0→1 use a twelve-month proxy: AOV × orders per customer per year × gross margin. | *Orders & AOV* and *Repeat-customer rate* queries, plus the cost sheet. |
| **LTV:CAC** | LTV divided by CAC. Below 1 loses money on every customer; 3 is the conventional healthy figure at scale. At 0→1 the sample is too small — read it as a direction, not a verdict. | Derived from the two above; state the window. |
| **Sell-through** | Units sold as a share of units received or made in a period. Whether the range and the buy were right. | Inventory levels (connector) against units sold via the *Top products* query. |
| **Return rate** | Orders or units returned as a share of orders sold, with the reason. | Shopify returns and refunds. `total_sales` is net of returns and `gross_sales` is not — say which a report uses. |
| **Repeat rate** | Share of customers in a window who have bought before. | The *Repeat-customer rate* query, as defined in `convention-shopify`. |
| **Email revenue share** | Revenue attributed to email and SMS as a share of the total. The owned channel, so a healthy DTC brand earns a meaningful share here. | *Channel mix* query by `order_referrer_source`, cross-checked with the email tool's attribution; state the model. |
| **AOV levers** | The moves that raise average order value: a free-shipping threshold just above current AOV, sets and bundles, a cross-sell at product page and cart, gift-with-purchase, tiered pricing. | AOV from the *Orders & AOV* query, before and after each lever, one at a time. |

Report against baseline and target, as `convention-shopify` requires; never as a bare number.

## The channel model

A channel earns its place by stage. The trap is opening every channel at 0→1 and diluting the one that matters.

| Channel | At 0→1 | At scale |
| -- | -- | -- |
| **DTC store** | The only channel that must exist. Where the brand, the customer list and the data live; every other channel feeds it. | Still the centre; margin and customer ownership are highest here. |
| **Marketplaces** (Etsy, Amazon, Not On The High Street) | Demand validation without paid spend for a maker brand; the cost is fees and a customer you do not own. Optional. | Incremental volume, kept secondary — margin and brand control both suffer. |
| **Social commerce** (Instagram, Pinterest, TikTok shop) | Discovery, not checkout. The content is the channel; a native shop earns its place once the catalogue is stable and a posting cadence feeds it. | Worth the checkout integration once a meaningful share of sessions originates there. |
| **Wholesale** | Not yet, unless a few aligned stockists buy credibility. Wholesale price is roughly half retail, so the margin has to carry it. | The volume lever for homeware once production allows; needs a price list, minimums and lead-time terms. |
| **Pop-up / physical** | Markets and pop-ups as research — watching people handle the product and hearing the objections is the cheapest discovery there is. | Seasonal pop-ups as brand moments, not a fixed store. |

## The fulfilment models

Three models are in play. The choice sets margin, lead time and the returns posture, so it is a strategy decision per range, not an ops detail.

| Model | Margin | Lead time | Returns |
| -- | -- | -- | -- |
| **Own stock** | Highest. Cash is tied up in inventory, so sell-through and the size of the buy matter. | Dispatch is the only wait. | Come back to you; restock or write off. Full control of packaging and unboxing. |
| **Print-on-demand via Printify** | Materially lower, and set by the provider's base cost rather than by you. No stock, no cash tied up. | Several days' production before shipping; state it on the product page. | The provider does not restock, so a return is written off; the returns policy must say so. Packaging is largely out of your hands — close the brand gap at steps 5 and 7. |
| **Maker-made small batch** | Set by the maker's price; usually between the two. | Weeks, and capacity-bound, so a launch is a dated drop — on-voice for BARA (real scheduled scarcity, never manufactured urgency). | To own stock. Provenance is a brand asset; say who made it at step 3. |

Mixed models in one store are normal; lead time and returns terms sit on the product page per product, not per store.

## The live instances

* **BARA / madebyneud** — UK ceramics and homeware, the store in `convention-shopify` (`www.madebyneud.com`), at 0→1. Own stock plus maker-made small batch, **self-fulfilled by the Owner personally** — no third-party fulfilment provider; a Welsh provider (Eirios) is a future decision tracked in MBN-298, not current state (Owner directive, APP-1535); the rebrand from madebyneud to BARA is in progress and the relaunch is the second run of `playbook-store-launch`. Brand instance in `convention-brand`.
* **Keroma** — apparel on print-on-demand via Printify, a fast launch; the first run of `playbook-store-launch`. Not yet in `convention-shopify`'s store table and does not yet have an operator; both are added as it comes online.

Each instance grows here as it trades and its numbers land.

## Read by

Read before: `skill-storefront-merchandising` (steps 2 and 3), `skill-lifecycle` (steps 5, 7 and 8), `skill-newsletter` (email revenue share, step 8), `skill-launch-campaign` (drops, discovery, the channel model), `skill-retail-strategy` (the channel and fulfilment models), `skill-unit-economics` (the KPI vocabulary, contribution margin above all), and `playbook-store-launch` (the chain as its checklist spine). `standard-storefront` sets the bar a storefront passes or fails; this convention supplies the model it is judged against.

## Quick checklist

- [ ] Work placed at a named step of the chain, with that step's one job stated?
- [ ] Every number reported in the vocabulary above, against baseline and target, with its query or sheet named?
- [ ] Margin stated as gross or contribution, never just "margin"?
- [ ] Channel chosen for the venture's stage, not opened because it exists?
- [ ] Fulfilment model named per product, with lead time and returns terms on the product page?