In October 2012, we launched a mobile wallet that let people pay with a phone.
The idea was early. The product still failed.
We called it MyWallet. I led the venture inside T-Mobile. My project record puts the investment at about €50 million and the delivery structure at roughly 200 people across eight companies.
We had banks, card networks, handset makers and other partners. We had working technology. We had a national brand and millions of customers.
We did not have a simple reason for enough people to change how they paid.
We built an ecosystem before we proved a habit
MyWallet launched nearly two years before Apple announced Apple Pay in September 2014. The title of this article is deliberately sharp, so let me make the distinction clear. We were building a mobile wallet before Apple Pay. We did not build Apple's product, and our architecture was different.
Our design placed payment credentials on a special NFC SIM. The official T-Mobile service terms show what that meant for a customer. They needed a supported NFC phone, the MyWallet application, an NFC SIM, data access and a compatible partner service. Some cards had to be installed through the bank and provisioned onto the phone or SIM.
Every dependency looked reasonable to the organization that owned it. Together, they created friction.
The customer had to want mobile payment enough to navigate handset compatibility, a SIM change, an application, a bank relationship and card installation. A plastic contactless card already worked.
We had designed a sophisticated route to a destination customers could reach more easily another way.
The first number was enough
By early January 2013, my project record showed about 5,500 activated users against a T-Mobile customer base of roughly 13 million. That is about 0.04 percent.
The figure should have changed the decision.
Instead, the scale of the effort made stopping harder. We had spent money, signed partners and built a large delivery machine. Each achievement became another reason to continue. We treated progress inside the project as evidence of demand outside it.
I was responsible for that choice.
This is the part of failure stories leaders often edit out. Market timing mattered. Technology standards mattered. Partner complexity mattered. The sponsor still had to interpret the evidence and decide whether more investment would improve adoption.
I kept believing that a more complete ecosystem would solve the problem. The evidence was already pointing at the customer journey.
Then the architecture moved
At the end of October 2013, Android 4.4 introduced Host Card Emulation. An Android application could emulate an NFC card without a provisioned secure element such as the one on a carrier SIM.
That did not make every mobile-payment problem disappear. It did weaken the strategic value of the architecture we had spent years coordinating.
Apple announced Apple Pay in September 2014 with availability planned for October. Apple controlled the device, operating system and user account. The announcement emphasized simple card setup and payment with Touch ID.
The contrast hurt because it exposed the real advantage. Apple made the path short.
Our technology worked. The experience asked the customer to manage our ecosystem.
Five decisions I would make differently
If I were leading the same venture today, I would make five choices before scaling it:
- Prove one repeated customer behavior. I would test whether a defined group pays more often with the wallet after the novelty fades.
- Measure the whole path. Awareness, activation, provisioning, first payment, repeat payment and failure recovery belong in one funnel. A large launch cannot compensate for a broken middle.
- Expand partners after adoption evidence. Each new bank or device adds coordination. I would add that complexity only when it removes a measured customer constraint.
- Treat architecture as an expiring bet. The technical design would have an evidence date and an explicit response to platform change. A working system can still lose its strategic reason to exist.
- Write the kill criteria before pride arrives. I would name the adoption threshold, deadline, decision owner and destination for released capital before the public launch.
These choices sound obvious after the failure. They were difficult inside a prestigious programme with strong partners and a future-facing story.
That is why kill criteria matter. They preserve judgment when effort, reputation and identity begin arguing for continuation.
The scar changed how I lead
I once believed that seeing the future early was a strategic advantage by itself.
MyWallet taught me to ask a harder question: can customers reach that future through the product we built today?
A brilliant concept can fail at the last metre when the surrounding system makes adoption harder. Partnerships increase credibility and coordination at the same time. The leader still has to decide whether the next investment will improve the customer path.
The €50 million figure is my account of the investment and still requires my final factual approval before publication. The lesson costs less to use.
Look at the project in your portfolio with the strongest technology story and the weakest adoption evidence. Trace the customer path. Set the date when continuation must earn its case again.
Your move.