Powering nearly one trillion digital interactions a year (calls, messages, chat, video and email), Twilio (TWLO) is the leading customer engagement platform used by hundreds of thousands of businesses globally to reach their customers. Twilio is expanding into new international markets with the goal to deliver an exceptional payment experience to new customers in every market. Twilio needed a payment processor equipped to handle international transactions at scale with high authorization rates to maximize revenue uplift, reduce issuer declines, and, in turn, reduce involuntary churn. For large businesses, even a single percent increase in authorization rates can deliver outsized revenue gains.
Twilio ran an A/B test with multiple major global payment processors monitoring authorization rates across card networks like Visa, Mastercard, American Express, Discover, Diners, and JCB. The test was set up to randomize routing by examining numerous transaction attributes like card type, card brand, card country, transaction size, and transaction type. After weeks of A/B testing, Twilio selected Stripe.
“We were seeing ~10% uplift overall in authorization rates,” said Chris Nasson, Director of Investor Relations & Treasury.
Twilio saw an increase in authorization rates by 10% compared to its previous provider. Here's a more in-depth look at Twilio's 10% uplift in authorization rates with Stripe, which it runs on AWS, as its payment processor:
+5.5% uplift from Stripe's Global Payments Infrastructure
With Stripe, Twilio is directly integrated with the major credit card providers—Visa, Mastercard, and American Express—and can process payments locally in its key markets like Japan, EU, and the Americas.
Nasson stated, "Our engineers don't need to worry about the nuances and the complexity of the Japanese market, which is by far one of the more challenging markets to process payments. Stripe makes our lives easier by reducing all that complexity in entering a new market."
+1% uplift from Adaptive Acceptance
Because Stripe is both an issuer and an acquirer, Stripe drives higher authorization rates through visibility and experimentation on issuer behavior on behalf of Twilio. If a decline does happen, Stripe's Adaptive Acceptance feature leverages machine learning to retry network declines in real time using optimized routing and ISO messaging, helping to improve authorization rates and recover revenue that may otherwise be lost. With Adaptive Acceptance, Twilio was able to mitigate issues with declined payments, which led to a 1% uplift over the last year.
"Stripe has a relentless focus on innovation. A great example of this is the machine learning model used to enhance authorization rates. We've realized roughly a 1% uplift in authorization rates just from Adaptive Acceptance. It requires zero effort from our side. It’s purely done on the engineering efforts of Stripe."
+2% uplift from Card Account Updater
Additionally, Stripe's Card Account Updater has yielded a 2% increase in revenue by intelligently updating the credentials of saved cards that have expired or been replaced.
+1.5% uplift from on-going consultation & support
On top of Stripe’s payment optimization products, Twilio opted for Stripe because of its ongoing consultative support. Stripe's team of payments optimization specialists analyzed Twilio’s transaction types and frequency to identify authorization rate optimizations by standardizing address verification treatment and maximizing local acquiring benefits.
“Stripe's Account Management team was really instrumental in ensuring our migration was on track. They offered new ideas for how we could better leverage the Stripe platform. Out of all of the different quantitative suggestions, we've received roughly 1.5% uplift to our revenue,” said Nasson. Stripe’s deep understanding of the nuances of authorization rates, as well as Twilio's product and goals, helped in identifying areas of improvement, which led to considerable revenue growth.
Stripe has a relentless focus on innovation. A great example of this is the machine learning model used to enhance authorization rates. We've realized roughly a 1% uplift in authorization rates just from Adaptive Acceptance. It requires zero effort from our side. It’s purely done on the engineering efforts of Stripe