The leadership team were excited about the possibilities of driving engagement and high user growth with the feed but the design team had concerns about the approach. We weren’t sure that users would always appreciate being highly social within the context of payments which could be sensitive and felt we were in danger of simply copying a competitor without understanding the pain points we were trying to solve for first.

If we were going to challenge the feed model effectively I felt we needed to come back with an alternate answer to the companies desire to achieve high organic user growth and increased engagement.

Whilst the wider team was having early discussions on licences, technology stacks and compliance I worked closely with another designer in my team to draw out a more detailed approach to how we could take the existing PagoFX app and learnings and move it into a more P2P driven space whilst supporting the existing international payments approach we had already built. We came up with an alternate approach to the feed model based on chat driven payments (like WhatsApp but focussed around payments) This was an area that we could see PayPal, Messenger and WhatsApp moving into already so it wasn’t entirely new but we wondered if it was a more natural way to be social in the payments space.

We brought several key innovations during this phase such as a simpler risk based ML and compliance model working with our peers in legal and product management which meant a less disruptive KYC flow (crucial for a more domestic P2P experience), the ability to both send and request payments, making use of open banking and most crucially of all a way to sending money to someone without them needing to have the app first so we could drive the network effect.

The main objective however was to challenge the feed model and change leadership hearts and minds around this which we successfully did after sharing the results of the testing and first hand feedback from users. We conducted a ‘within subjects test’ of both prototypes (all participants saw both prototypes but we alternated the order) to remove bias and get significant data on which one worked best with a small sample size.

The preference was clear – the chat based model was more natural and easily understood than the feed approach. Many participants had concerns about what was and wasn’t shared on the feed and they stated many examples where they wouldn’t want certain family or friends to see the details of payments they had made (birthday’s, embarrassment etc).

To that end we created an extended chat prototype that included the sign up process / bringing someone new into the network to collect a bueno payment, more detail around the ML and KYC process, sending to and requesting from an individual and also sending to a group.