Home / Case Studies / SodaStream
SodaStream · Subscribe & Save
~30% CR uplift
How a structured A/B methodology lifted SodaStream's subscription conversion by ~30%
The challenge
SodaStream's Subscribe & Save flow asks for three consequential decisions before it asks for money — cylinder type, quantity, and delivery frequency — and each one is a place to lose a subscriber. The first is the worst: a shopper who does not know whether their machine takes a Quick Connect or a screw-in cylinder cannot proceed at all, which is why the step carries a "not sure?" helper. Quantity and frequency then have to be made for the visitor rather than by them, via a most-popular default and a saving stated against the alternative. Because subscriptions compound into lifetime value, a gain here is worth far more than the same gain on a one-time purchase.
The approach
Evidence first: the flow was analyzed step by step to locate where prospects hesitated and dropped — the cylinder-type decision, the quantity default, the frequency commitment — so hypotheses came from observed behavior rather than opinions.
Structured A/B methodology: every candidate change was written as a falsifiable hypothesis with a pre-registered primary metric before launch — no post-hoc goalpost moves.
Sequenced testing: tests were ordered so each result informed the next, letting winning patterns compound instead of running disconnected one-off experiments.
Guardrails throughout: changes had to win without degrading order economics.
The result
The program compounded to close to a 30% conversion-rate uplift on the subscription flow — not from one silver-bullet test, but from a sequence of evidence-ranked experiments building on each other.
The takeaway
Subscription funnels reward methodology over inspiration: a disciplined test sequence on three decision steps beat any single redesign.
Your turn
Same method. Your funnel.
The audit finds your version of these leaks — priced in recoverable revenue, with the test plan attached.