Case studiesFurniture retailClient system
Canvas + Loft
A handwritten thank-you for every order, approved before it is sent
Canvas + Loft wanted every customer to get a real card in the post, not just an email. Cards turns Shopify orders into thank-you cards, puts each one in front of a manager, and sends the approved ones through a handwriting service or a print-and-post service.
- Client
- Canvas + Loft
- Product
- Cards
- Industry
- Furniture retail
5
Shopify webhooks feeding new orders into the queue
2
ways to send: handwritten, or printed and posted
3
decisions per card: approve, flag or skip
The problem
Follow-up after a sale depended on who remembered.
The store sold well, but what happened after the sale was down to memory. Some customers got a thank-you, most heard nothing, and nobody could say who had been sent what. Writing and posting cards by hand for every order was never going to happen.
- Most customers heard nothing after their sofa arrived
- No record of who had already been thanked
- Handwriting and posting every card by hand took too long to keep up
How it fits together
Shopify orders
- Card sequences
- Manager approval
- Handwritten cards
- Printed and posted cards
- Send log
What it does
Approval queue
Every order lands in a queue where a manager checks the message, the recipient and the address, then approves, flags or skips the card. Nothing is sent without that review.
Sequence builder
Decides which customers get which card, and when. Sequences can have several steps, several message variants and hold windows, such as waiting 14 days after delivery.
Send and log
Approved cards go out through a handwriting API or a print-and-post API, and every send is logged. Deduplication makes sure a customer is never sent the same card twice.
Inside Cards
Card sequence
A handwritten thank-you a day after fulfilment, then a referral postcard eight weeks later. Each step knows how many times a customer has reached it.
Real screens from the live system. Names, figures and personal details are blurred.
The impact
Every customer thanked, without anyone writing a card by hand.
Cards is feature-complete and in final setup before the first real cards go out. Measured items describe the system as built. Projected items are our estimate, with the assumption stated.
What changes when a thank-you card is a queue instead of a good intention.
Assumes about 300 orders a month with 80% approved by the manager.
Projected figures. They are our estimate for a business of this size, with the assumptions written under each number below. We replace them with measured results as clients confirm them.
Measured
Memory → queue
for deciding who gets a thank-you card
Every Shopify order now arrives in the approval queue by webhook instead of relying on someone to remember.
Projected
~4 to 5 min
saved per card versus writing and posting by hand
Assumes writing, addressing and posting one card by hand takes 4 to 5 minutes. In Cards, approving one takes seconds.
Projected
~0% → 100%
of approved orders that get a card
Assumes almost no customers received a card before, and that every order a manager approves is sent once the system is live.
Under the hood