Framework to figure out what features to launch for your product

Imagine, you are running a commerce platform. In the early days of running it, you decide to launch a chat feature to generate more engagement on the platform.


Does it make sense? While the most important metric is the number of transactions, you have digressed by trying to improve engagement on the platform.


Often while building the product, we digress from the core problem we are trying to solve.


Why does this happen?


Adding new features often becomes the solution to all our problems. While running Reach (my first start-up), we did 3 redesigns in 3 months. Our wild young mind kept on floating ideas and we kept building features after feature confusing the user.


Framework to follow:


  1. Create a Feature document and attach one key metric that you expect to increase with launch of this feature.

  2. Plan all your activities around that.

  3. Circle back to this document every 2 weeks to make sure you haven’t digressed from your core problem.

Another recent example, while I was consulting for a company, we got overly consumed with improving numbers that were vanity metrics (easily manipulated numbers). They didn’t really add any meaningful value to our customers, but gave the false impression of improvement.


When we should have been sticking to providing value to our customers, we ended up chasing false metrics.


Which begs the question: how do you (or does one) stay true to launching features that provide value to customers?


Here’s a framework that I love to follow:


1. Do user interviews almost daily.


Taking user interviews is an art. Idea is to look for patterns, without any bias.

For example, while working on a sleep music app, one common issue that my users were complaining about was that they had to get up during their sleep to stop the music. Here, we Introduced a timer feature that allows the users to put a timer to the sleep music. This thus adds value to the customer and is not some fancy feature launched with no objective. Here’s a document I have made on how to do user research.While I was working at ixigo (India’s largest meta travel company), my job was to create a smart AI bot who could resolve the queries of the customers. The first task I did was to talk to our customer support team daily for 30 min. This helped me to understand the variety of queries our customers keep asking and bucket them into different verticals. This exercise gave me the confidence on what our customer wanted and a clear direction to detail out the product feature.


2. Wherever in doubt, put feedback loops into your product. This will help with more data and better analysis.


While working as a product manager, we often try to relate user behaviours with our own behaviour of using an app. This can get tricky.You are not the customer — keep reminding yourself this.

The key is to have data to back any of the hypotheses you make.


What kind of feedback loops am I talking about?


After observing data, you see a lot of users exiting the app within the first hour of app launch.


Feedback loop — Initiate a pop-up when users are exiting the app asking what went wrong? Is the App buggy? Is this not what you were looking for


B. User validating product’s a-ha feature: On Tinder, when you get a match, it asks you to rate the app. That’s a feedback loop.


Feedback loop — Initiate a pop-up right after the a-ha moment, asking the user if they liked the product. The chances of getting a response to the feedback will be higher.


You are confused whether the feature is providing any value to the customer.

Feedback loop — On exiting the feature, show a pop-up asking for feedback, if you liked the session? Did the feature work well? Try to put the issues that they are likely to face and validate those.


3. Talk to your customer support team DAILY. They are the ones responding to the queries of your customers. You need to know the problems your customers are facing.


While I was working at ixigo (India’s largest meta travel company), my job was to create a smart AI bot who could resolve the queries of the customers. The first task I did was to talk to our customer support team daily for 30 min. This helped me to understand the variety of queries our customers keep asking and bucket them into different verticals. This exercise gave me the confidence on what our customer wanted and a clear direction to detail out the product feature.


This framework has helped me stay close to my customers and keep building zero to one features which are centred around solving the pain problems that my customers face.



I publish weekly blogs where I share everything I learn.
Subscribe below to stay updated! 👇🏽

Thanks for submitting!