• Stripe’s machine learning model, Adaptive Acceptance, created 1.24% of the total uplift leading to $1M+ in additional revenue.
  
• Card Account Updater (CAU) generated 2.76% of the uplift, resulting in $2M in revenue.
  

Zapier, the leading automation and workflow software for businesses, supports more than 7 million users worldwide, connecting thousands of technologies. Prior to Stripe, Zapier’s payment processor delivered poor authorization rates and didn’t provide clear reporting with transaction-level details, making it impossible for Zapier’s payments team to identify processing issues and improve authorization rates. So, Zapier began a search for an improved payments platform offering higher authorization rates, a selection of local payment options to support its global customer base, and transparent reporting to manage interchange rates and fees and ensure no money was being left on the table.

Zapier ran an A/B test with its previous payments processor and Stripe. The test was designed to randomize routing to eliminate bias among the experiment cohorts. Zapier analyzed results across several transaction properties, including card type, card brand, transaction size, and more.

“During the evaluation, we ran all of our failed payment charges from our previous payment processor through Stripe—25% of those originally failed payments that would have otherwise been lost were accepted on Stripe,“ said Ryan Roccon, controller at Zapier.

With Stripe, Zapier had access to intuitive, data-rich reporting tools, helping its finance team view interchange rates, declines, and refunds to understand the details of its acceptance rates. Over the course of the test, Zapier saw significantly higher auth rates compared to its previous processor, making the decision to switch to Stripe easy.

Integrating Stripe only took a couple of weeks. “We’ve saved hundreds of hours of engineering time by integrating with Stripe—clear API docs and Stripe’s support team got us up and running quickly, even with a complicated global integration,” Roccon said.

He added, “Stripe continues to keep its foot on the gas, investing in machine learning and other advanced automation to ensure they provide the industry’s best acceptance rates. Since changing processors, we have seen a 4% uplift in auth rates which has resulted in an additional $3M in revenue to our business that we wouldn’t have had without Stripe."

Adaptive Acceptance created 1.24% of the total uplift leading to $1M+ in additional revenue

Because Stripe is both an issuer and an acquirer, it has greater visibility into issuer and network behavior, which helps Zapier increase its authorization rates. Adaptive Acceptance uses machine learning to optimize authorization messages. Stripe’s machine learning models leverage Stripe’s extensive historical data, such as transaction type, issuer, and merchant type, from processing hundreds of billions in payments annually. Stripe’s data scientists and engineers continuously improve these models to help merchants accept as many legitimate transactions as possible, generating additional revenue.

Card Account Updater (CAU) generated 2.76% of the uplift, resulting in $2M in revenue

For customers, manually updating information on saved cards that have expired or been replaced can be tedious and time-consuming and result in lost revenue for your business. Stripe works with card networks to automatically update saved card details whenever a customer receives a new card (e.g., replacing an expired, lost, or stolen card). This allows your customers to continue using your service without interruption, reduces the need for you to collect new card details whenever a card is replaced, and decreases the chances of declines. For Zapier, CAU resulted in a 2.76% uplift in auth rates.