Why this matters
A paid first order rarely covers the cost of acquiring the customer (CACCost to acquire one new customer.). The repeat purchase and loyalty guide shows why the second order is where a store's economics turn, and the e-commerce email and SMS flows guide maps the messages that carry the work. The AARRR view for e-commerce sales adds pricing the first order and reading retention on each category's own cycle.
The merchant and its customers
Saffron Row Spices is an invented online spice merchant selling jars, cheaper refill pouches, recipe kits and a starter collection. Ads win first orders steadily; very few customers return.
Saffron Row Spices: customers, orders and flows
| Measure | Value |
|---|---|
| Cost to acquire a new customer | £19 |
| First order: average value / contribution after product, packing, postage and fees | £26 / £10 |
| Repeat order: average value / contribution | £22 / £9 |
| Customers placing a second order within 90 days | 11% |
| Repeat orders per customer in the first 12 months | 0.3 |
| First product bought | Second order within 90 days | Median gap to second order |
|---|---|---|
| Starter collection | 19% | 70 days |
| Recipe kit | 15% | 28 days |
| Single jars and blends | 7% | 55 days |
Current flows: three welcome emails, each repeating a 15% sign-up code; a cart flow with a discount in the first message; a review request 3 days after dispatch.
Contacts: 14,000 email subscribers, 6,500 of them customers. SMS consent has been collected at checkout for four months, from 900 customers.
Survey comments: "didn't know you sold refills", "ran out and bought some at the supermarket", "couldn't remember the shop's name".
Mapping the lifecycle
Find the paths from first purchase to second before writing any message. Recipe kits come back fastest, the starter collection most often, single jars least often. The comments say customers do not know refills exist and forget the store by the time a jar runs out. So the post-purchase flow points at refills and the next recipe, and replenishment is timed to each product's gap.
Task
- Price a customer's first year against acquisition cost, and the repeat orders needed to break even.
- Design the post-purchase, replenishment and win-back flows: trigger, timing, messages, channel and exit.
- Say what you would change in the welcome and cart flows.
- Propose one loyalty mechanic, the uplift it needs to pay for itself and how you would measure it.
- Name the repeat purchase metrics to track, and how often.
Worked answer
Saffron Row: the customer today
First-year contribution per acquired customer: £10 + 0.3 × £9 = £12.70, against £19 to acquire. Each customer loses about £6.30 in the first year.
Repeat orders needed to break even: (£19 − £10) ÷ £9 = 1.0 per acquired customer, against 0.3 today.
By first product, using the 90-day rates as a rough guide to the year: a starter collection buyer is more than twice as likely to return as a single jar buyer (19% against 7%), so the same £19 spent winning a starter collection buyer is closer to paying back.
Flows alone cannot close that gap, so acquisition should also lead with the starter collection and recipe kits, and lifetime value (LTVTotal value a customer brings over time.) is read by first product.
Saffron Row: the flow plan
| Flow | Trigger and timing | Messages | Channel | Exit |
|---|---|---|---|---|
| Post-purchase | Order placed | Dispatch note with storage tips; on delivery, a recipe for what was bought; day 10 after delivery, review request; day 18, refill pouches for the jars bought and a matching recipe kit | Email; delivery-day note by SMS where consented | Second order placed |
| Replenishment | Days since order, by first product: recipe kit day 24, jars and blends day 48, starter collection day 62 | A refill basket built from the last order, one click to buy; if no order after 5 days, a one-line reminder with a re-order link | Email, then SMS for consented customers | Order placed |
| Win-back | About twice the usual gap: kit buyers day 56, jar buyers day 110, starter buyers day 140; repeat buyers on twice their own gap | What is new since their last order and what goes with it; a week later, an incentive only for customers with two or more orders | Email only | Order placed, or sequence ends and the contact moves to sunset |
Why these choices. Post-purchase answers the comments: refills exist, and the store's name appears at first use. Replenishment lands before the median gap, before the supermarket wins. SMS carries only the short re-order nudge, to consented customers. Win-back opens with a reason, and sizes any incentive to past value.
Welcome and cart. The sign-up code goes out once with its expiry, and the welcome stops at the first order. The cart discount moves to the last message, for first-time customers only.
The loyalty mechanic. A store credit card in the first parcel, for a second order within the replenishment window. It targets the second order directly and carries no points liability. Watch average order value (AOVAverage revenue per order.) on credit orders in case baskets shrink to the credit.
Saffron Row: what the credit must achieve
Credit: £3 on a second order within 60 days. Cost per first-time customer = new second-order rate × £3. Gain = (new rate − 11%) × £9.
The credit pays for itself on the second order alone when new rate × 3 = (new rate − 0.11) × 9, which gives a new rate of 16.5%. Below that, it only pays if those customers go on to order again.
A holdout of 10% of first-time customers receives no card, so the difference in second-order rate between the two groups is the credit's effect.
Metrics. Monthly: second-order rate by first product and channel; time between orders; each flow's conversion; win-back reactivation; the holdout difference; unsubscribes and SMS opt-outs. Quarterly: contribution per customer at fixed ages against acquisition cost, by cohort.
Common mistakes
- One delay for every product. Kit buyers need a prompt long before jar buyers; the gap comes from order data.
- Leading win-back with a discount. It trains customers to lapse.
- SMS without its own consent. Email consent does not cover it.
- Judging the credit on redemptions. Customers who would have returned anyway redeem too.
Next, take the e-commerce plan capstone or return to the intermediate e-commerce level.
