From an idea to a global card product: How we launched the TransferWise debit Mastercard

Engineering

In 2018 we launched our TransferWise debit Mastercard — moving from online-only money remittance to our first physical product. We thought, if one can hold, send and receive money without hidden fees, it should be possible to spend it that way too. And a physical payment card felt like a natural step towards creating a borderless ecosystem for international use cases in finance.

Fast-forward a year and a half and with the help of a project team, loads of research and proof of concept we had a physical card product. We’ve launched the card in 32 countries, had over one million active cards and grown from a small team into a tribe of seven teams. It’s been quite a journey!

This blog is all about that journey — from customer-centric development to tackling new pain points and growing a product team into a product tribe.

Building a card product was different to anything we’d done before and we had new considerations to think about but what remained paramount was for us to understand how to take what our core product brought to our customers and translate that into a physical product.

TransferWise Borderless ecosystem

The main value proposition of the card is that it lets customers spend anywhere in the world at the real exchange rate. Once we’d begun the process, we started to understand the challenges our customers face with international money transfers, like opening bank accounts in a new country. To eliminate the problem, we launched local account details meaning customers can have account details in five currencies without physically being in the country (GBP, EUR, AUD, USD and NZD). This allows them to receive and use money as a local bank account holder. Within the borderless account, customers can hold and exchange funds instantly in 54 different currencies.

And, by continuing to get to grips with new customer needs and wants, we’ve introduced special, value-adding features to the card too. For example, if a customer holds multiple different currencies and spends money with the card, the system automatically spends in the currency with the best conversion rate at the time, saving them money which is a big part of our mission.

When everything feels important, it can be tricky to prioritise what’s going to be most valuable for our customers especially when building a product from scratch.

There were essential features like Google and Apple Pay that would take time to implement, multiple opportunities to take the card into new markets as well as making sure we were providing our customers with the best overall user experience. Alongside our priorities we also had a long wish-list of features from our customers…

So how did we cut out the noise and get moving?

We began by listening to our customers to get a handle on their needs and pain-points through a variety of channels:

  • NPS comments: With eight million customers using the product, we’ve automated how we analyse NPS feedback and categorise keywords into themes helping us get everything we can from the data we get from our customers. We then look into these themes to understand the biggest pain-points or areas for improvement.
  • User experience research: We have a dedicated user experience research team that simulates usage situations and harvests feedback from them. Members of product engineering teams are present in user testing sessions. Our user testing teams have developed strategies and tooling for these sessions but as the whole product team is invited, our product engineers and managers can ask any questions they like. These sessions give us insights into the understanding, behaviour and emotions our customers have when using our product.
  • Feedback from customers: We have a team of ‘Customer Support Champions’ for the card. They collect and compile information from customer contacts about the product, and every two weeks we get together to understand what users have experienced. It could be things like technical issues on our end or external issues like ATMs not functioning in a particular country. Knowing about a problem helps us stay up-to-date with our customers and the market.

Our analyst Triinu, senior engineer Tiddo and product manager Karam brainstorming

Once we’ve done all that, we’re able to prioritise and can define the impact of each proposed solution on customers. We aim to be as data-driven as possible and it’s important that we’re able to quantify the impact and forecast how we’re going to move key performance indicators (KPIs) with each feature or project.

The data allows us to pick the projects with the highest customer impact, and then we divide the work into sprints and work in an agile framework to deliver the new features. Once we’ve implemented a new feature, we measure the results and compare them to our forecast. If the results don’t align with our expectations, we go back and try to understand why, and how we can estimate better next time.

New features are great but the strategy and roadmap for product engineering depends heavily on the product performance dynamics. The fate of the product adoption would cause a corresponding roadmap. Teams should better be always mentally ready to see their rocket failing to fly. If it does, it’s sometimes best to just nervelessly shut the project down and reorganise people to do something else. Perhaps another rocket. Or make some flying one fly faster.

In our case, we’ve been enjoying really decent growth so we haven’t had to prioritise any dedicated product growth projects or ponder what missing features would make our customers start liking the product more. Instead, our primary focus so far has been on the technical progression — it’s a big luck and luxury which made it possible to raise the stakes and magnify our technical ambition. We started with a proof of concept debit card solution to validate the market fit. Then we replicated the initial MVP in 30+ countries on 4 continents. To get the card to more potential hands, we also launched cards for businesses and added token payments including Apple, Google, Garmin and SamsungPays.

Today our card product is quite lightweight and simple. Many parts of our UI are still considered an MVP. It’s because at that stage we didn’t jump immediately into working on product’s bells and whistles, but rather completely re-architected both the whole tech and the team that builds or operates it. We focused on designing a highly modular, scheme and partner agnostic flexible platform for the future. An architecture that’s quick and easy to evolve and can effortlessly scale. Using the racing analogy, I like thinking that we’re building a car for 24 hours of Le Mans in a way that it could be successfully pushed to the limit on track for straight 300 hours or more. Engineers on the team of course love building it. And soon we’ll zero in on the product enhancements for our customers again and the tech underneath should make us move really really fast.

Good product health means an obligation for tech excellence. And tech perfection creates a vital soil for further rapid product development.


I like old-school whiteboarding. With tribe located in 3 offices around the world, we need cameras to keep our cross-site teammates involved

TransferWise has lengthy experience in the payment card industry. Card pay-ins is a popular method of uploading money into TransferWise. The pay-in product is owned by our “Cards and Pay-in platform team”. Once we had built and established the card acceptance worldwide, the team started to look for new challenges within the industry. One of the chosen challenges was creating our own payment card.

With a temporary sub-team of two we’d validated the idea with market research and proof of concept. Once the idea was looking good, we created our “Plastic team” that focused on building the product full-time. As we’ve grown the product, we’re now called the “Plastic Tribe”, dividing up into smaller product teams for clearer focus.

The team that launched our card in first markets in 2018

The tribe consists of product teams with their own focus areas as well as regional product teams alongside Product Analysts, Operations, Fraud and Compliance Operations and Customer Support Champions who support the teams by bringing expertise from their specialisms.

Plastic Tribe structure

And, what unites the teams is the mission of the tribe: “to give our customers a reliable, easy to use and eventually, free way to spend from borderless accounts anywhere in the world”. Each quarter the tribe sets tribe level objectives that will get us closer to the mission and each team in the tribe builds their projects and OKRs (Objective and Key Result) to make the objectives translate into tangible deliverables. Every sub-teams’ deliverables and objectives filter down from the mission and this helps us to stay aligned.

Our team and tribe structure are constantly evolving according to customer needs. There are elements in the structure that allow us to move fast and focus on the most impactful projects so we’re always thinking about our customer first.

  • With our autonomous team structure, we create new teams when they’re needed to bring focus to a specific area or customer group. At the moment our Experience team is a single team focusing on improving the customer experience of card users. But, with our Business debit card launch it’s clear business customers have different needs to individual ones, so to build the best experience for each customer group the Experience team could be split to cover both areas.
  • Division into regional teams helps us optimise for different markets. To ensure we’re responsive in all customer time zones, we have local talent in the market that understands local customs, customer behaviour and culture to optimise the product locally. However, we don’t have pure regional card product development teams focusing just on geographical localisation. Why? Our engineers want to focus on developing the core technology and, as a side responsibility, make it work well in their home market too. So as an example, our Singapore-based engineers primarily own the sensitive cardholder data protection part of the system while taking care of card developments related to APAC region (Cards in Singapore, Australia, New Zealand and new markets in the region)
  • Support from Product Analytics, Customer Support, Fraud and Compliance help us move fast. Specialists in these areas help us focus on the right things at the right time: Product Analysts and Customer Support Champions help us focus on customer pain-points and needs, whilst Fraud and Compliance ensures compliance and mitigated fraud risk to us and most importantly our card holding customers.

We’ve come a long way since we had the idea to take an idea and build it into a physical card product but there’s so much more to do. We’ve got big plans and are really excited to see what the future brings. Watch this space!

P.S. Interested in working with us? We’re hiring!

Check out our open Engineering roles here.