Rally: one-click checkout for e-commerce brands
The backend for a checkout that replaces the e-commerce platform’s own, with one-click buying for returning shoppers and post-purchase offers in the same order.
- E-commerce and payments
- Portland, Oregon (remote)
- 2020-2024
- revenue from post-purchase offers for one merchant
- +38%
- a day through Rally for a single merchant
- $30k
- commerce platforms: BigCommerce, Swell, Salesforce
- 3
- raised on this backend, from seed to Series A
- $18M
The problem
Brands on BigCommerce, Swell and Salesforce Commerce Cloud were stuck with their platform’s built-in checkout. Returning shoppers had no one-click buying.
Any upsell after payment became a second order with a second card charge, which shoppers abandon and merchants hate fulfilling.
What we built
The backend for a checkout that replaces the platform’s own and stays in sync with it both ways. Cards are vaulted and tokenized, shoppers verify with SMS codes, and orders and stock update in the merchant’s store in real time.
Post-purchase offers land in the same order as the original cart, which wasn’t possible before without splitting it. The hardest work happened where shoppers never see it:
- Speed: we cut checkout start-up time down to the millisecond.
- Gradual rollout: a merchant could send, say, 10% of checkouts to Rally, and a shopper who landed on Rally kept getting Rally even after leaving and coming back.
- Customer data: as soon as a shopper entered an email, we pulled their existing customer record from the platform and merged it with what they had already typed.
Results
- +38% revenue for one merchant from post-purchase offers.
- Up to $30k a day through Rally for a single merchant.
- Brands like PopSockets checked out through Rally.
- Live on Salesforce AppExchange since 2023.
- The backend carried the company from founding through a $6M seed and a $12M Series A.
The smart layer
Merchants built their own offer funnels from simple rules, for example: “Cart has X and Y but not Z and is over $80: offer A; if accepted, offer B, otherwise C.” A machine-learning service in Python studied how funnels performed and suggested new ones to each merchant. Every payment went through two integrated fraud-screening services.