Zerodha: Warn people, even when it costs you trades.
Zerodha added warnings that slow customers down before risky trades, even though every trade earns a broker money. Explore when protecting users from a bad decision builds the business instead of costing it.
See the idea in one picture ↓- first nudge
- 2020
- drop in penny-stock trading
- 50%+
- Nudge v2 announced
- 2026
Figures as reported by the company.
Independent educational analysis. Scenarios, people, numbers and illustrations on this page are original PMcademy practice material, not company research results or an endorsement.
A speed bump, not a wall
- 1A normal order
Most trades go straight through with nothing in the way.
- 2A warning first
A flagged penny stock shows a warning. Most people stop, and anyone can still proceed.
- 3Limits you set
Nudge v2, announced in 2026, adds hard limits that customers set for themselves.
Animated illustration. Names, numbers and people are fictional teaching material.
Your revenue and your users’ interests can pull apart.
A risky buy is one tap away
A pause, a warning, then your choice
Read the analysis
Imagine your product earns a small fee every time someone acts: a trade, a bet, a purchase. More actions mean more revenue this month. Some of those actions hurt the people taking them, and people who get hurt leave or stop trusting you. The product question is not only how to increase activity. It is which activity you want more of, and which you should slow down on purpose.
Add a speed bump, not a wall.
Illustrative data and simplified behaviour. This model does not represent a live company system or measured company results.
Read the analysis
Use the model below to compare three set-ups. In the first, a risky order goes through like any other. In the second, a warning explains the risk, and the person can still proceed. In the third, the person sets their own limits in a calm moment, and the product holds them to it later. Notice how much of the choice stays with the user in each case. The numbers are illustrative.
A first-time investor does not know what they do not know.
“A friend said this stock will double. Should I just buy?”
Read the analysis
Our fictional learner, Kavya, opened a trading account last month. A message tells her a small stock is about to double. She has never heard of price manipulation. A plain buy button makes the risky choice as easy as the safe one. Ask which of your users are most likely to make a costly mistake, and what they would have wanted to know a moment before making it.
A drop in activity can be the result you wanted.
Zerodha says penny-stock trading on Kite fell by more than half after the warnings.
The baseline period or what customers did instead.
- Warnings heeded vs overridden
- Losses avoided
- Customers active a year later
Read the analysis
Zerodha says penny-stock trading on its platform fell by more than half after the warnings, and presents that as a success. That is the company’s own account, with no baseline period published. In your project, decide in advance which fall in activity would count as success, and pair it with signals of long-term health such as complaints, losses avoided or customers still active a year later.
Some customers will call it nannying.
- Customers protected
- Long-term trust
- Fewer costly mistakes
- Fewer trades now
- Some call it nannying
- Warnings can become noise
Read the analysis
Warnings protect people who did not know the risk, and they annoy people who did. Too many warnings become noise that everyone clicks past. Blocking an action outright protects more but takes the decision away from adults who may have good reasons. Decide where a warning is enough, where a limit the user sets is better, and where, if ever, the product should refuse.
How far should the product step in?
Your fictional team runs a buy-now-pay-later app. Some customers take on more instalments than they can repay. You can show a warning before the fourth purchase, let customers set their own monthly limit, or block new purchases after a missed payment. Choose your approach, explain what it costs the business in the short term, and name the long-term signal that would justify it.
Choose a next move
Leave with a guardrail plan.
Your artifact
Write a guardrail plan: the risky actions in your product, the guardrail for each, the message people see, and the short-term and long-term metrics it moves.
Critique my project artifact using only the evidence I provide. Write a guardrail plan: the risky actions in your product, the guardrail for each, the message people see, and the short-term and long-term metrics it moves. Separate observed facts, assumptions, proposed tests and missing evidence. Identify a failure mode and a cheap check. Do not invent company results or user research. Ask for my artifact before starting.
Read the analysis
List the actions in your product that can harm the people taking them. For each, choose a guardrail: information, a warning they can override, a limit they set themselves, or a hard stop. Write the message they would see, the short-term metric it will lower and the long-term metric it should raise. Add one way to stop warnings from becoming noise.