Distribution First

I used to be in the "build a better mousetrap" camp. reMail's product was better email search for your iPhone. Very constrained. Most users found reMail through tech blogs or by word of mouth. It was a cool product, but growth was decidedly non-exponential.

When you look at the most successful startups of the last 5 years, one common property is that they built not just a polished product, but a great distribution strategy that was a natural fit. In some cases, it was a viral marketing scheme like Dropbox's - sign up your friend, get free storage. In others, it was that the creators had identified a segment of potential users that would immediately adopt the product - think about how Facebook forcibly created an account for all Harvard undergrads on day one. Gilt came to own New York by spamming women in their core age and income group.

In a way, coming up with distribution ideas is more fun than coming up with startup ideas, partially because it is more constrained: You have to invent a scheme that's not too expensive to operate, socially acceptable, and targeted to the demographic that would use your product. I encourage you to think distribution first.

Four Types of Mobile Apps

I like this blog post by Chris Dixon that talks about the four kinds of mobile apps: Time wasters, core utilities, episodic utilities, and notification-driven apps. I think these four relatively narrow definitions cover about close to 100% of what people actually do with their phones.

Phone makers often talk about how with today's technology, you have a supercomputer in your pocket. Yet no one wants a supercomputer in their pocket - none of the app categories require large-scale processing. The games that need fancy 3D graphics are often not that popular - they take too long to load and are too complex for the small screen.

It's the simple, vertical apps that solve well-defined immediate problems that work best.

Much more left to do

"Probably the biggest secret is [...] that there are many important secrets left. This used to be a convention forty or fifty years ago. Everyone believed that there was much more left to do. But generally speaking, we no longer believe that. It’s become a secret again."


-- Peter Thiel in Stanford's CS183

reMail now builds on iOS 5.x

I know some of you folks are still following this blog from the days of when reMail was open sourced.

While I haven't added a lot of features to reMail since then, I've been trying to keep it up-to-date and buildable. The changes that happened in iOS 5.0 and XCode 4.2 broke the build a while ago, and a variety of downloadable zip files sprung up with forks of reMail that build on iOS 5.x.

The good news is that I just pushed a patch that builds reMail on iOS 5.x flawlessly. Many thanks to Matt Ronge who put this patch together. He helped a lot with reMail even back in the day: reMail uses the MailCore framework he built.

Android Location Best Practices

Android location is complex: You have to register with different location providers, listen for their updates, derive the phone's location - all this while you're watching for battery consumption.

I've found myself pointing a lot of people to this excellent tutorial by Android's very own Reto Meier: Deep Dive into Location on the Android developers blog.