Digital Lucky Money revamp
Digital Lucky Money needed mechanics that matched how people actually prepare and give lucky money in Tet Holiday. The revamp changed how random splitting worked, removing the social friction that had quietly undermined sender confidence for two seasons.


ROLEProduct designer
TEAMTechcombank • Payment
TIMELINENov 2023 - Jan 2024





Background

With two Tet seasons behind it, the feature was due for a refresh. The team was asked to improve the experience and update the visuals ahead of the next season. Research changed the scope of that brief.


Uncovering insights
As-is experience
We conducted a usability test with 6 participants on the as-is flow, reviewed the previous year's data, and gathered direct feedback from internal bank users who had used the feature.
Random splitting, intended to reduce the mental work, introduced social friction instead. With no control on the range, the system could produce a minimum amount that felt disrespectfully low, which senders only find out after the packets were opened by recipients. They lost confidence in creating and giving out packets again.






Behavioral data from both previous years reinforced the stakes: 80% of senders chose random for direct giving. The friction wasn't in a rarely-used edge case. It was in the dominant behavior.



Reflection on the traditional lucky money experience
Perceived value comes from the sense of good fortune, not just the amount. Banknotes with lucky serial numbers are an established norm in traditional lucky money giving: patterns like 1111, 8888, or 6868 are sought out specifically. The digital equivalent had an opportunity to reflect that norm.





Design challengeThe problem wasn't randomness itself. It was that the mechanic gave senders no way to make it feel safe or intentional.

How might we give senders confidence in a feature built on randomness?





Solutions: Rethink Randomness



Auto random, but under controlThe setup screen now shows the minimum amount as an editable field, with an optional maximum. Senders set the floor, and the system handles the rest. The social risk of an unexpectedly low amount is removed before the packets go out.



Split into lucky-looking numbersThe random amounts are distributed into patterns that carry cultural weight. The sense of luck is built into the numbers themselves, not just the act of giving. The patterns generate automatically. Senders set the budget, the numbers take care of themselves.



22,222

68,686

79,797

67,890

...



and progressive disclosure for senders who want granular controlAn edit view is available to adjust the distribution.







In-person group giving, default to randomWith the minimum under control and amounts in lucky-number patterns, random splitting was no longer a social risk. It became the default for in-person QR giving, where over 80% of senders were already choosing it.





Before we simplified
The early exploration offered denomination sets, each generating amounts in patterns tied to specific lucky numbers. Senders could shuffle between sets until one felt right.

It worked culturally. But it reintroduced a choice at exactly the point the experience should feel effortless, so it was simplified to a single toggle, with the system handling the rest.






What came next


The core advantage of digital lucky money is removing the need to prepare in advance: no exchanging for new notes, no running short at the last moment.
The template
would let senders open the app, review a pre-filled setup, make edits if needed, and Confirm under a minute.
It was added to the development backlog for the following Tet season.






My reflection

Lucky money is a deeply social experience.
Understanding what produces fear as well as confidence in the giver, and delight in the moment of receiving, turned out to be the real design problem underneath the usability issues.

Defaults are design decisions.
Changing what's pre-selected changed behavior without adding steps or complexity. The flow sequence stayed the same; what changed was the logic inside each step. In a feature users already know, defaults carry more weight than new features do.



Thank you for reading!





NEXT
Techcombank Design System