logs

We are launching this week. There’s a lot of get done, and the success of our product depends heavily on how this launch goes.

Day 235: Monday, December 12, 2022

We worked throughout the weekend to finish up all the remaining major coding tasks for our AI chat interface. Planning/copy for all the launch messaging is mostly finished, but there’s still some minor things left to do (create videos, clean up wording, coordinate friends for launch, etc.).

Our plan at the start of the day was to launch by tomorrow—this is technically possible, but it was starting to feel quite rushed (plus Kevin and I are still remote). We decided to push the launch back to Thursday (2 days), which sounds like a lot but it’s actually just one day as we’re launching on Product Hunt on Thursday midnight. This means that right after I land back in the Bay Area, we have a few hours to clean things up, and then we’ll launch. Hopefully I won’t be too tired.

At the start of the day, our TODOs for launch were:

Brought back the “retry” mechanic within our chat interface.
Pricing infra/page
Set up onboarding
Write email to current users with list of feature changes
Finish writing all the copies for marketing on platforms
Finish demos/videos
Send out messages to friends/prominent individuals to help share/like our launch
Landing page with new demos
Spruce up Discord
Product roadmap

Day 236: Tuesday, December 13, 2022

We moved our launch date yesterday from today to Thursday (wow that’s a lot of days in one sentence). Nothing new to add here, as we just worked on our list of TODOs. Also added a bunch of various bug fixes/UX improvements to the list. I spent a long time planning, recording, and editing our demo videos for launch—definitely harder than I initially anticipated.

Day 237: Wednesday, December 14, 2022

Quite a tiring day for me. Woke up early for my flight (got 3 hours of sleep), but I got a large coffee at the airport so that I could get some work done at the airport/on the plane. I was literally running on fumes at this point, but there was so much stuff I needed to get done that I had no other option.

I had rough drafts for most of my marketing copies, but I still wasn’t really happy with how it looked. Anyone who’s written a shitty rough draft for a school essay before finally cranking out an acceptable final draft right before the due date knows this feeling. I spent essentially my entire time at the airport/plane rewriting everything, sending emails to high profile figures in the generative AI space asking for them to share our launch, and creating good demos.

After I landed, we still had a lot of stuff we needed to do. I took a few minutes to take a shower/eat dinner, and then it was back to grinding away in preparation of our launch. John essentially built out an entire onboarding pipeline that night while Tae worked on improving our landing page copy. I kept working on our marketing materials and coordinated my friends for our Product Hunt launch at midnight.

Our launch technically occurred at midnight (so passed over to Thursday), but it feels like the same day so I’ll include it in this log. Our Product Hunt launch went well by some metrics—because I had so many friends ready to upvote as soon as I posted it, we got like 30-40 upvotes right away. However, I think Product Hunt is actually good at detecting this sort of fake upvote behavior because we didn’t actually make it to the front page all night. I went to sleep pretty disappointed and also a little sad about the lack of success—this was compounded by the fact that I was the person in charge of launch itself, so I felt responsible for its success/failure.

Day 237: Thursday, December 15, 2022

I woke up really early (7 am) to go online and post our remaining launch posts on Twitter/LinkedIn—I also went ahead and sent an email to all our current users announcing our launch. I went to our Neo Slack and asked the community to help reshare/like our launch posts, and also sent a link to my Tweet to several higher profile individuals on Twitter. Around the same time, our Product Hunt post finally made it onto the front page (lfg), and we continued to get a steady stream of upvotes throughout the day. I thought it was pretty amusing how Mem AI released a similar product on Product Hunt, but we managed to stay above them in the rankings nearly the entire day (after we finally got featured). Almost immediately after posting on Twitter/LinkedIn, we started getting a steady stream of signups—much more than our Product Hunt launch. Kevin and I did have to put out a mini fire when there was an issue with our onboarding (there was a pretty major bug), but we resolved it early in the morning before Tae/John woke up.

Things really kicked off at around 8:41AM when the higher profile individuals (Linus Lee and Ben Tossell) retweeted my launch post. Hundreds of signups started pouring in, and it was really exciting to finally see our hard work pay off—Tae and John actually woke up due to the constant buzzing of Slack notifications (we get sent a message every time a new user signs up). A lot of the dread I’d been feeling about if our launch was a failure finally vanished, and I felt like I’d drank four cups of coffee (in reality, I was running on like 8 hours of sleep total across two days). We passed 500 users at 10AM, 800 users at 1PM, and 1000 users at 3:22PM PST.

Overall, the launch was a huge success, and I am super happy/excited about the future.

Day 238: Friday, December 16, 2022

Users were still coming in this morning—I spent quite a bit of time responding to users. More generally, we pushed out lots of bug fixes/UX improvements from user feedback—things are really picking up after launch.

We’ve also been garnering quite a bit of VC interest—as I’ve stated earlier, our philosophy has been to raise only when we have a need for money. This is because raising funding takes so much time, and you have greater leverage when you actually have a plan/specific uses for said money. Typically, you also raise later because you’ll be bigger/be able to demand better terms. That being said, however, generative AI is quite different in that we are in a clear hype cycle, similar to crypto (it is different from crypto in the sense that there are very clear use cases of AI that weren’t possible without). We don’t want to exit the hype cycle without having raised, and we also want to capture the momentum/excitement from our launch—hence why we are now open to conversations with VCs.