Muhammad Usman

Muhammad Usman

Co-Founder & COO

Badges

Gone streaking
Gone streaking

Forums

Has an incumbent ever eaten your differentiator mid-build?

MyFitnessPal acquired Cal AI in March. Anyone building around AI photo logging in that category woke up to find the feature they were going to lead with now belonged to the market leader, attached to 280 million users.

I was one of them. Didn t stop, repositioned.

But I don t know yet whether that s the right instinct. Three options exist when this happens compete on execution, reposition around something the incumbent can t retrofit quickly, or stop and the advice online is entirely survivorship stories from people who picked one and got lucky.

If this has happened to you: which did you pick, and did it work? And is a feature-level differentiator always borrowed time against a company that size, or does execution actually hold a wedge open long enough to matter?

Does anyone actually vibecode native iOS? Everything I read assumes React.

Just finished an iPhone app native Swift, Apple Watch target, voice logging through App Intents. Built most of it with AI tooling driven off markdown specs.

What struck me: almost every vibecoding post, tool demo and tutorial assumes a web stack. React, Next, Tailwind. The models are visibly better at it more training data, instant feedback loop, errors you can read in a browser console.

Native iOS is a different experience. Xcode errors are worse to feed back in, SwiftUI changes enough between versions that the model confidently writes last year s API, and there s no quick way to show it what the screen actually looks like.

Still faster than writing it by hand, but the gap between works for web and works for Swift is wider than the tooling marketing suggests.

A month of daily content, zero signups. What actually got you your first 100 users?

Posted short-form video every day for a month across four platforms. Thousands of views, zero signups. Not one.

What I think I got wrong: assuming owned channels compound. They don t when you start with no audience you re feeding an algorithm that has nobody to show you to.

What s worked better since: direct outreach, and communities where the problem already gets discussed.

So for your first 100 users, what actually moved the number? And has content marketing ever worked for anyone from a standing start, or is every success story survivorship bias?

View more