decision one
A QR on the table, not an app.
The diner scans and lands straight on the tip. No download, no signup, no account. Faster than finding change ever was.
✕Rejected: a tipping app. A download kills an eight second impulse.
Tippd · a true story
The night a thank you
had nowhere to go.
One dinner, one brilliant waiter, and no way to tip him. This is what I did about it.
The problem
The tip died with cash.
UPI pays the bill. Nobody pays the waiter.
I owned
All of it. Sole designer.
strategy · research · design · code · payments · onboarding
I built
One QR, three wins.
tip to the person · praise to Google · bad nights to the owner
Week one
₹1,900 · 65+ · 3.9★
tips to staff · new reviews · up from 3.8★
1
It started with a bill.
A waiter looked after us all evening. Water refilled before we asked. He remembered what my friend could not eat.
👛I reached for my wallet. No cash. There never is anymore.
🧾I asked them to add a tip to the bill. They said no. Their billing cannot do it.
📱So I paid by UPI. The restaurant got everything. He got nothing.
HOTEL ANNAPOORNA
Table 12 · 9:02 pm
Paneer 65240
Butter naan × 4180
Veg biryani310
Fresh lime soda × 3150
TOTAL₹880
UPIPAID ✓
Tip for Sandeep. . . . . .
Nowhere to put it
2
So I went back, and listened.
Not to build. Just to ask. Tap a card to hear more of what they said.
The surprise: I walked in believing tips would win the owners over. They nodded politely. They only leaned forward when Google reviews entered the conversation. That one realisation reshaped half the product.
3
Three walls, every single time.
Tap a tile to turn it over.
4
So Tippd became three decisions.
Everything else on the screens serves one of these.
decision two
A face before an amount.
You are not paying a bill, you are thanking Sandeep. His photo comes first, the money comes second. The kitchen can be thanked too.
✕Rejected: one pot for the restaurant. Gratitude has a face.
decision three
Money moves when you hold, not tap.
Holding for one second makes it deliberate. No accidental taps with money. Try it, it works.
✕Rejected: one tap pay. Too easy to fire by accident.
press and hold
decision four
Give the owner a reason to say yes.
Tips alone do not convince an owner to put a stand on every table. Reviews do. After the tip, one question. Good nights get a nudge to Google. Bad nights go quietly to the owner, before they become a one star review.
✕Rejected: a public review prompt for everyone. Bad nights deserve privacy, and owners deserve the warning first.
And the ones I got wrong first
5
Then I designed it. Twice.
The pilot below runs v1. This is v2, the redesign it taught me, running live right here. Drag the amount, hold to send, flip the phone.
1Scan the QRno app, no signup
2Pick the persongratitude has a face
3Set the amountdrag up, be generous
4Hold to senddeliberate, never accidental
5It landshis phone buzzes, not the till
6One questionpraise public, problems private
live prototype
6
Then I stopped designing, and shipped.
🎨Designedin Figma
💻Builtevery screen, all the code
💳PaymentsUPI, integrated by me
🤝Pilotedin person, on the floor
📈Reportedweekly, to the owner
7
Then the numbers started talking.
The first pilot: Inti Bhojanam, Chennai. They started with 804 Google reviews at 3.8 stars. One week of Tippd on the tables.
3.8★ → 3.9★
the restaurant's Google rating, one week later
₹1,900tips, week one
65+new reviews
0negatives public
1bad night caught
Good morning! Weekly progress of Tippd 😊
3.8 → 3.9⭐ in one week 🔥
One 2⭐ night was flagged in time. It reached you, not Google.
23 Jan 2026, 9:04 amOne restaurant, one week. Every number from the report we sent the owner.
8
And now it lives on real tables.
Chennai. Real tables, real waiters, real tips.
photo from the restaurant, coming here