I am trying to compile all of Nikita Bier’s wisdom on twitter. The problem is he shitposts as lot. I have gone through all of his tweets in 2023 and below is the result. I will hopefully cover 2021/22 also. If you find this useful, please tell me. That will give me the motivation to go through earlier years as well.
Read about Nikita’s story [2]
Mega thread: [4]
Gas app walkthrough: https://x.com/MacBudkowski/status/1811018742190485586
Insights/interesting stuff:
- Aim to be #1 on App store quickly
- The main thing about going viral is to have people talking and sharing your product for you. A great way to do that is leave out certain information to build up curiosity and speculation [1]
- feed into people’s deep desires [1]
- For every social app I’ve ever built, the number of invitations sent per user drops 20% for every additional year of age—from 13 years old to 18. So if you build for adults, expect to pay to acquire every user with ads [3]
- on PMF: Instrumentation is so hard to get right that data will often deceive you, particularly with questions of product-market fit. So, how do you know if you have product-market fit? It will be so abundantly obvious that data will be the last thing that is giving you conviction. [8]
- Nikita is relentless: “Do you A/B test? No, I just ship the app—and if it's not ranked in the App Store, I change it until it is.”
- On building teams: “If I was putting together a team to make a new viral app, the expectations would be simple: You will work for 10 hours a day for 7 days a week and put all other life plans on hold for 4 months. In exchange, you will own double digit equity, draw no salary, but 3 meals a day will be provided. If it doesn’t start working by the end of 4 months, we disband. If it does work, you will make more in a few months than you have in your whole life.”
- Product Architect (layout, flows, hierarchy, marketing, roadmap) Technical Designer (prototypes, animations, visual design) iOS Engineer Backend Engineer User Support
- Clientside SMS convert at around 30%. 5 invites x 0.3 = 1.5 K-Factor
- Everytime I build a high brow product, it flops harder than I could ever predict. Everytime I build something trashy, it grows beyond all expectation. Never forget that humans are animals.
- I’ve set up a new constraint for the products I make:
- Development time cannot exceed 3 weeks.
- Revenue must be a minimum of $3 million in 90 days.
I’ve counted over 20 apps in the App Store that have achieved this, 3 that were launched in the last 3 months.
- Social apps is one of the few areas where people on the outside think "oh I have an idea and that looks easy" so it's inundated with first timers trying their hand at it. But in reality, it took me 5 years to figure out how to effectively onboard people and form a social graph.
- The number of people we text peaks at 25
- You can learn more about your users by putting a giant Give Feedback button in your app (with live chat) than if you spent $1 million on professional “user research”. If you aren’t there for them when they have a complaint, they go to Twitter and the App Store reviews. Catch negative sentiment before it snowballs and control the narrative.
- How to decide when to give up: 3-5 launch attempts to eliminate potential confounding variables associated with the Go To Market method.
- The components that have the most engineering scope tend to be wholly reusable—such as the friendfinder & invite system. The fastest way to arrive at an app that resonates is to run rapid experiments by only changing the interaction model. Without properly building the reusable components, you will often walk away from tests wondering if it would work if users had sufficient friends. Eliminating that confounding variable will allow you to pivot with conviction.
- Related: if it's a social app, you should always have a base set of network features so you can rule out confounding variables (i.e., friendfinder, contact sync, invites, friends-of-friends)
- Who to build for? You should build for whichever demographic meets these three criteria:
- Has a growing social graph
- Sees their friends daily
- You can personally relate to
For No. 3, I just have a conversation with my thirteen year old self and make whatever I would've wanted.
- Every app I’ve built I’ve quietly tested in the American South and Midwest for many months.Testing on urban coastal audiences distorts your data and draws premature awareness to your experiment. If it works in Middle America, it will work everywhere.
- Valuations: [9]
- About FB: One comment from a parent increased churn by 10x for teens. It was on a death march once it expanded outside of schools.
- About simplicity: There is a tendency for app designers to create layers & subgroups to deal with complexity: Mastodon attempts this with usernames—which have 2 parts. For every part of your app that you fragment, expect to increase your app’s overall probability of failure by 50%. Users don’t have the patience to learn about the subworlds of your community. They are more motivated to churn than to understand. Early products already have a limited inventory of content. When you fragment things, average engagement per post takes a hit, which is the key metric to track the health of your app. Subgroups should be completely off the table for new products and the absolute last resort for mature products. Sometimes your urge to fragment your content is your own anxiety to keep things organized: resist this urge.
- We tend to overcomplicate what it means to have great product sense. To put it simply: you can look at a screen and intuitively predict the percent of users who will convert to the next screen within a 10% margin of error. This skill is just so plainly valuable in all walks of business beyond just design: in sales, recruiting, content development—really, anything. This is why a company's product organization is so often conflated with the company's leadership.
- When designing a consumer product, you should consider every tap by a user to be a miracle. The motivation to stop using a new app will always be stronger than to use it. To signup & complete a profile on Gas, we got it down to 15 taps total—with no keyboard required at any stage. Needing a user to type with a keyboard at any stage to become activated is asking for failure.
- From the outside, a breakout social app seems like it would just work on the concept alone—without grinding out optimizations. In practice, the difference between exponential decline vs. growth (0.99 vs. 1.01 K-factor) is often an extra invite button or ranking contacts correctly1542827
- not trying to become snapchat and be valued at multi-billions or focused on retention. He’s really just looking to build something exciting, go viral, and cash out. [1]
- Nikita is pro sharing contact list [5]
- If you’re a social app that requires contact sharing, don’t market to techies. They’ll tear you apart for it. Instead market to everyone outside of the bubble and you’ll improve your chances. [6]
- I always geofence California and New York out of the app and launch there last. Not worth the distraction when you’re scaling up.
- On pitching: 99% of people are so bad about talking about their product that they need to practice on morons in a low-anxiety context. to form their narrative arc [7]
Strategies:
- To get people to notice and become curious about the app, they would first create instagram accounts with the high school and brand name together. For example, @gas.georgiahigh. They would make the account private, then follow everyone associated with that high school. As the students would request to follow back so they can see the profile, they would not accept anyone. Second, at 3:00pm on the day before launch, when all the teens got out of school at the same time, they accept everyone at the same time. So every student is getting the same notification at the same time, showing each other and talking about it, to then go to the profile directing them to download gas to “see who likes you.” [1]
- When deciding on the authentication system for a social app, you might be inclined to allow users to pick any method—like Google, Facebook, Phone, Email. But if you truly want to grow & get friend density, you should only allow one method—and have that method be on the same “protocol” as your distribution & friending system. For example: • Phone Number Auth should be paired with SMS Invitations and Contact-based Friendfinding • Twitter Auth should be paired with Share to Twitter and Import Followers. When you allow multiple methods, it leads to a corrupted social graph where users can’t find each other and the app won’t be able to properly rank people that users should invite.
Questions:
- Geofencing - how can it be done? why?
- easy to be #1 on app store
- to make it “restricted access” building the hype and excitement
References:
To do later:
- https://kickstartsidehustle.com/ find a 10 page case study on Gas/nikita here