This was a pretty fun week.
Readings
Read Masters of Doom, a book about the creators of Doom and Quake :). 9/10 book and a must read for anyone who just wants to hear a good story. It literally does not have a boring page. The author does such an amazing job with pacing which is so hard to do when you’re covering the lives of real people. How long do you spend on their childhood stories? When do you start talking about their first breakout hit? I’ve been reading many more of these “story” type non-fiction books lately. I learn a lot from them. From Master of Doom, apart from just getting to hear a good story, I had a key learning: when building some new, you need to have people around you who inspire you and push your ideas to the next level. Else? Well. You end up stuck in your own head.
This resonated w/ me. Especially during quarantine, Zip has been getting lonely.
Right after finishing the book, I went out to find 4 people I wanted to work on with. A day later it was done. I’m working with two teachers and two developers (all part-time) whom I like, respect, and can help push me/Zip further. You can follow the two developers on Twitter here and here.
I also read a paper titled 3D Photography using Context-aware Layered Depth Inpainting where the authors massively improved inpainting in 3D. Basically, you take a 2D image + a depth map, run it through their algorithm, and then explore other areas of the image (originally not in the image) in a 3D setting. I pulled their code and ran it to see it for myself. Here’s the original pic w/ me + my mom.

And here’s the result. The swing is insane, the way it understand the foreground, the background, and makes it feel like the image was shot in 3D.
I also did it with the homies from Naruto and the effect is much more pronounced.

Look at Sakura!!! The way it goes around her head in insaneeeeeeeee.
Also, read this cool blog post titled A (possible) solution to COVID-19, highly recommend reading it. Basically the idea is to create a vaccine using strains of COVID-19 found in the wild vs perfecting one in the lab which can take years. The author talks about how plenty of people have harmless versions of COVID-19 where there is a natural deletion in the gene that causes the life-threatening effects.
Ah shit I just popped a pimple on my neck that’s disgusting IT BURNS ARRRGGGGHH.
Zip
Last week I clearly laid the problem: Zip has people showing up at the door pretty easily. I just need to press a button on my Facebook Ads dashboard and the people come flooding in. I can spend $100 and get around a 1000 people pretty easily. Great!
This week has two focuses:
Did people like the thing? Why or why not?
Will people come back to the thing?
The first thing I did this week was rehaul my email system. I was using Mailchimp but Mailchimp is fucking garbage when you want to customize your flows. So, I set up Sendgrid which seemed to have a really Go library. Now, I send emails:
After they signup for the email list, I send a “welcome” email giving them access to their lessons. With SendGrid these send nearly immediately vs Mailchimp which took 30-60 seconds.
A reminder email 48 hours after people signup for the email list and didn’t check out their portal.
A survey email 1 hour after they check out the portal.
An email telling users when they have a new week’s worth of lessons available in their portal. That’s right! Multiple weeks :). This took 2 days to setup. I kept running into random edge cases w/ existing prod users.

I also fixed all my analytics w/ Amplitude so I can clearly track my funnels which I wasn’t doing an awesome job at before. I even got Sengrid working w/ my analytics so I can even track email opens/clicks all in one place. My user’s entire journey can be analyzed in one place.
It’s around a 40% conversion rate right now from facebook ad -> email -> checking out app.

I’ve been getting tons of awesome feedback from the survey I setup to.

So, things are looking good. People seem to like the thing. Cool. I thought it was shit. But, this is why we ship to users instead of being stuck in our heads :). Seems it does just enough where people find value in it. I got 550 signups in the last 5 days from Facebook Ads, of those 550 I’ve also had 24 people tell me directly they love the product. That’s 4%. It costs me $55 to drive those 550 people. So it costs me $0.43 to get someone who loves my product right now. That’s an interesting #. A better product would drive acquisition cost down and perhaps get people to even pay for the damn thing…probably?
I still don’t know if people will come back to the thing every week. I just shipped the “week updates” yesterday, so, not enough data yet. This morning I sent 88 ”new-weeks” emails and 17 checked out their new lessons which is 20%. Cool! On a Saturday, that’s not bad. I have 17 WAUs confirmed thus far :).
What I really wanna see is how much it costs me to retain a WAU. Then, I can scale up spend on Facebook to $100-$300 and see what happens. Ideally, I’m charging for Zip by end of month. I think by next week I’ll have a better idea about all this since I’ll have 2+ weeks of usage data.
So there’s the update! I still don’t feel good about Zip current state (don’t think I ever will be) but I feel like I’m on to something. I feel like I’ve been on to something since December. But, right now the feeling is stronger than ever.