Building at @Google often means building for scale is non-negotiable. Lots of people (including me when I joined a little over a year ago) think that means lots of technical difficulties. However, I have realized that the tech is actually easy. It's people that are hard. (1/n)

Building a niche product is massively different from building one for the masses, especially when it's a paid product. (2/n)

I have been a part of countless brainstorming sessions where software engineers pitch new ideas for the product which sound amazing to me and I am sure would sound great to quite a few of you in the Software/Tech industry. (3/n)

Their mistake is that they think "I am an accurate representation of my product's audience". Let me assure you, you are not ! (4/n)

I am sure surfacing ML based insights on product usage sounds like a super fun project and it seems like a great feature to have, however, is it really going to get your product to go from 20M users to a 100M ? (5/n)

When you're dealing with a TAM of 500M+ users, you need to think deeper to move the needle in any significant manner. And here's a thought that might make you hate me. At that scale, all your intuitions are wrong ! (6/n)

You and your ideas are heavily influenced by the biases you have developed in your life. At that scale, people having the same set of biases as you dwindle very fast. (7/n)

The only thing that keeps the product running and growing at that scale is DATA. That is why most teams at Google (at least the ones that have the budget) do extensive user research before deciding the theme/roadmap of their product. (8/n)

And honestly, data never lies. Even if it does not make sense to you, following that data is probably the best thing you can do for your product. (9/n)

The next time you think why doesn't product 'X' with millions of users have feature 'Y', try asking a diverse set of 50+ users , "would you pay more money for feature 'Y' ? ". The answers might help you a lot in thinking about building products. (10/10)