25%
of annual revenue
through email and SMS
Case studyRetention · Bareback Footwear
Thirteen flows, and
one of them is a business.
A quarter of annual revenue arrives through a channel that costs nothing per send. It was built from an empty list over ten years, alongside the paid media that fed it, and rebuilt from scratch in 2026 when the old flows stopped earning their keep.
100k+
subscribers,
grown from zero
13
live flows,
in four groups
39.96%
of customers come back,
averaging 1.7 orders
Paid media rents attention. The list owns it.
Every pound of Meta and Google spend buys one visit. The only way that visit keeps paying is if the customer joins something: a list, a loyalty balance, a reason to hear from the brand again. So from the first year the job was not just to sell a boot but to keep the person who bought it.
Ten years on, the list is the most valuable asset the brand owns that is not stock. It does not need to be re-bought every month, it does not move with an auction, and it is the channel that turns a first order into a second. Nearly 40% of customers come back.
Acquisition & recovery
Welcome seriesJoined the list
Abandoned checkoutCheckout started, not finished
Browse abandonmentViewed a boot, did not buy
Back in stockAsked to be told
After the order
Post-purchase, split on first vs repeatOrder placed
Fit check, day sixBranch of the above
Win-backGone quiet
The loyalty engine
Voucher claimedReward redeemed
Friend's referral codeReferral invite sent
Referral creditFriend has bought
Birthday creditDate based
Fit & service
International orderShipping abroad
Wide-fit guidanceBought a wide-fit boot
Eight emails, two customers.
The post-purchase flow opens with one question: has this person bought before? A first-time buyer and a repeat customer get the same four beats at the same moments, with different words, because they need different reassurance.
First orderRepeat customer
2 hoursA personal note from the founderWelcome back, from the person who packs the boots
Day 6How do they feel? The exchange promise leadsHow do they feel? The fit question leads
Day 11 to 13Looking after your new bootsLooking after your new boots
Day 41 to 43Room for one more pair? Early accessRoom for one more pair? Early access
The day-six email is the hinge. It asks about heel, toes, calf and instep once the boots have actually been worn, offers a free exchange for sixty days, and its answers rewrite the fit levels on the product page. That is how the retention programme feeds the returns programme. The returns case study picks that thread up.
Built, not bought.
Off-the-shelf loyalty apps assume every product behaves the same way. Boots do not: they get exchanged, they come in seconds, and a referral is worth a different amount depending on who is being referred. So the scheme was written against Shopify and Klaviyo directly, with the rules living in the brand's own logic and Klaviyo doing what it is good at, which is sending.
1
A friend's code
A customer shares a £10 code. The friend gets it by email straight away. Five days later the flow checks whether the friend has ordered in the last thirty days. If they have, it stops. If not, one nudge, then it leaves them alone.
2
Credit on completion
When the friend buys, the referrer is told their £10 credit is on its way. The money is released by the scheme, not the email, so the rules about when it becomes spendable sit in one place.
3
Credit on redemption
Spend a reward and a plain transactional email confirms the voucher and its value. No marketing in it. It is the receipt for something the customer earned.
4
A birthday
£5 in the balance on the day, from a date the customer chose to share. Small, expected, and the kind of thing people mention to a friend.
Rebuilt from scratch in 2026.
Flows decay. Copy that was right three years ago reads stale, products it points at have been superseded, and the segments underneath drift. In the summer of 2026 the thirteen flows were rebuilt from a blank canvas rather than patched: new copy, new timing, the first-versus-repeat split added to the post-purchase flow, and the loyalty engine wired in as four flows of its own.
The old flows stayed live while the new ones bedded in, so nothing went dark on the way through. That is the unglamorous part of retention: not the welcome email, but the discipline to rebuild it before it stops earning.
Trigger on behaviour, not the calendar
A flow that fires because of something the customer did is read. A newsletter that fires because it is Tuesday is deleted.
Split on history
A first-time buyer and a fifth-time buyer need different words at the same moment. One question at the top of the flow is enough to give them that.
Own the rules
Write the loyalty logic against the platforms you already run. It is less work than it sounds and it will match how the product is actually sold.