UPI keeps setting records while its average payment gets smaller. Credit cards win the large purchases; UPI wins the daily habit. This project measures the split and designs a product that lets a premium issuer take part in small-ticket spending.
Billions
₹ per transaction = monthly value ÷ monthly volume
| Month 2026 | UPI value (₹ L cr) | Card spend (₹ L cr) | UPI ÷ cards |
|---|
Persona: Rohan, 34, consultant, Gurugram. He holds a premium travel card on a global network for flights and hotels. Every day he also makes ten or more small payments by scanning QR codes from his savings account. Those payments earn no rewards, don't use his credit limit and never show up for his card issuer.
For a premium issuer, Rohan's daily habit is invisible. The issuer sees a few large transactions a month and loses the frequency that builds loyalty.
| Element | Design choice | Why |
|---|---|---|
| Companion RuPay credit | A virtual RuPay credit card issued on the same credit limit as the primary card, linked to UPI in two taps. | RuPay is the only credit network allowed on UPI. |
| Small-spend rewards | Flat points on QR payments, capped each month, that pool into the primary card's rewards. | Builds frequency without unlimited reward cost. |
| One statement | Primary and companion spends on one bill, one due date, one app view. | No extra complexity for the customer. |
| Spend controls | Per-transaction and monthly caps for QR payments, set by the customer. | Reduces fraud exposure and over-spending. |
As a premium cardholder, I want my daily QR payments to earn the same rewards as my card, so that I don't have to choose between convenience and value.
The proposal assumes premium cardholders want rewards on small QR payments. That is the riskiest belief, so I would test it first, with 8 people who hold a premium card and use UPI daily.
| I believe | How I'll test it | What would change my mind |
|---|---|---|
| Premium cardholders would move daily QR spends to credit for rewards | Interviews plus a fake-door test: an in-app "Link to UPI" button that counts taps | Fewer than 1 in 10 people who see the button tap it |
| A second card number is acceptable if it lives only in the UPI app | Show two concept screens; ask which they'd turn on | Most reject any second card |
| Spend caps reduce fraud worry more than they annoy | Ask people to set their own cap in a prototype | Most leave caps off or call them a hassle |
Research status: not yet run.
Monthly UPI volume and value for January–October 2026 come from NPCI releases as reported in the press. I derived average ticket size and the ratio of UPI value to RBI-reported card spend. The data (each row with its source link) and the script are in this folder; the SQL version is in the data workbench.