Subscription Pricing
Date: 2026-08-16
Trading a lower price for a repeat commitment. The discount is only affordable because the second and subsequent orders carry no acquisition cost — so the whole decision is an LTV arithmetic problem, and the retention rate is the input that decides it.
What it is
Subscription pricing offers a recurring purchase at a discount to the one-off price, in exchange for commitment.
For a consumable retailer it’s the strongest available lever on Customer Lifetime Value — not because the customer spends more per order, but because they keep ordering without being reacquired.
Why the discount is affordable
The one-off customer costs £22.50 to acquire and orders 1.6 times. The subscriber costs the same to acquire and orders as long as they stay.
one-off customer
1.6 orders × £15.00 contribution = £24.00
− CAC £22.50 = £1.50 over their life
subscriber at 10% off
contribution per order £15.00 − £5.00 = £10.00
average 6 orders before churning = £60.00
− CAC £22.50 = £37.50
Twenty-five times the lifetime contribution, on a lower per-order margin. The discount is bought back by order count — LTV to CAC Ratio, Payback Period.
The whole case rests on that “6 orders”. Get it wrong and the arithmetic inverts:
average orders before churn lifetime contribution net of CAC
2 −£2.50
3 £7.50
6 £37.50
12 £97.50
At 2 orders the subscription loses money against a one-off customer. So the number that decides subscription pricing is retention, not price — Retention and Churn, Cohort Analysis.
Setting the discount
Work back from the break-even order count rather than picking a round number:
discount contribution/order orders to beat one-off (£24.00 lifetime)
0% £15.00 1.6
10% £10.00 2.4
15% £7.50 3.2
20% £5.00 4.8
30% £0.00 never
At a £15.00 contribution on a £50 order, a 30% discount takes contribution to zero — every subscriber then loses money forever, at scale, quietly. That’s the specific failure this table exists to prevent.
Note the shape: the discount comes off revenue but the costs don’t move, so a 10% price cut is a 33% contribution cut. Discounts on thin-margin goods bite far harder than they read — Discount Impact on Margin.
Structures
| Structure | Mechanism | Suits |
|---|---|---|
| Subscribe and save | Standing discount on each order | Consumables, the retail default |
| Prepaid term | Pay upfront for N deliveries | Cash flow, strongest retention |
| Membership fee | Flat fee unlocks pricing or shipping | Where breadth beats one product |
| Tiered commitment | Deeper discount for longer term | Where churn is early and predictable |
Prepaid retains best because the churn decision has already been made and paid for. It also brings cash forward, which matters more than it sounds for a business holding stock.
Membership decouples the fee from any single product, and converts the pricing question into a usage question: how many orders before the fee pays back for the customer, and for you.
Where subscriptions leak
The pricing is rarely what kills them:
- Cadence mismatch. Delivering faster than the customer consumes is the largest single cause of cancellation. Make interval easy to change — an easy skip beats a hard cancel
- The unskippable order. One delivery the customer didn’t want cancels the subscription permanently
- Price rises mid-subscription. The subscriber’s reference price is what they signed up at, and a rise is felt as a breach — Price Anchoring
- Failed payments. Cards expire; a meaningful share of “churn” is involuntary and recoverable with retries and pre-expiry prompts — Involuntary Churn and Dunning
- No visible ongoing value. After a few orders the discount stops feeling like a benefit and starts feeling like the normal price
Measuring
- Contribution per subscriber cohort over time, not conversion to subscription. A subscription programme that converts well and churns at order three is worse than no programme — Cohort Analysis
- Cannibalisation. Subscribers who’d have bought at full price anyway are pure discount cost. The comparison is against a matched non-subscriber cohort, not against the average — Incrementality Testing
- Voluntary versus involuntary churn, separately. They have completely different fixes
- Interval changes and skips as a retention signal. A customer who skips is retained; a customer who cancels isn’t. Skip rate rising is good news being logged as bad