1. Hook (Opening Slide)
The fastest #1 on Producthunt ever (24h effort). Here is how it started πΉ
2. Context/Background
One late night doing random research on perplexity:
"Is there a lib out there, that can gen commit msg based on diff with local AI?"
nothing interesting
"doesn't llama etc. require big models etc?"
nothing interesting
"Can I use Apple Intelligence local model?"
"The "secret" backdoor to accessing these models from the terminal is the Shortcuts app"
3. The Journey (2-4 slides)
- Check if anyone else knows about this secret? Nope. Nothing. Only partial solutions. β
- Figure out a good product around this. With low UX friction. 100+ md research files.
- Figure out the "global maxima" around what could be built. Where are the limits?
- Make it not-boring.
- can you create an image. Miss Minutes from loki. but as a devtool character?
- maybe like a cat?
- maybe like a black cat is the mascot.
- that cat is nice. just focus on that cat. can you show me it in another pose.
- Can you show me just the head.
- make it look at the viewer β
- Build out the landingpage.
- Download counter (Social proof)
- Github star counter (Its opensource after all)
- 2 step installation. Shortcut + curl in terminal β (low friction 10 sec)
4. The Results
- Launched it on Monday. Spamed some frenz. After 2h. One of the most prolific PH hunters had hunted it.
- What just happened? When do we launch? Its up to you Zac Zuo (the hunter) he suggest Tuesday, the fierciest day of all. Mmmokay. lets roll with it. Easy come easy go.
- Tuesday. Wakes up. Nothing. Its been moved to Thursday by the PH team. Alright. At least it wasnt put in the not featured bin.
- Zac Zuo: Do what you can. Improve the hunt. Make a video or smth.
- 2 more days of improving the product. π Thanks for that! 5am On thursday. t minus32 h to launch. Fix all bugs + Make a launch video.
- Wednesday: Email from PH. No mention of featured status, but tmrw is your big day, dont cheat jadadjada. promising π
- Ai voice over β Ken burns zoom in on the app (iMovie) β lofi-beat background music β Show install + key features β
- 8am (featured on PH π), 9am (shoots up 70 upvotes straight out of the gates leading the race), 10am. im still slepping 11 am. still sleeping. 11.15am. check my ph ios app. the grace period is still hiding postiions, checks hunted.space. nr1. whoaa.
- alright. lots of comments, engagment, downloads, some treats. A YC company in our rareview π
- the hours go by. its a 24h race. But we hold of. And win by ~100 votes. around 4-500 votes total. 3-4h before the race is over, we still lead by 80 or so. I call it a night. The contendors doesnt seem to be pushing too hard, so it will be fine. and the next day, we did secure the nr1.
5. Key Takeaways
- PH featured status is down to: Creativity, Usefullness, Innovation, Craft. (but in the AI age, innovation is favoured, because the other 3 are now much easier to achive)
- So from the begining we asked. Has this been done before? Because PH will find out if it has. And then you dont have inovation. your a me2, a copycat etc.
- So if you finds something that is usefull, that hasnt been donebeofre. bingo, eurka. yu jsut found 100$ on the street. So start with that. competitor / market research.
- Being influential on PH helps a little (3.5k followers). But it doesnt matter in terms of if you get featured or not. Innovation esp, then nail the 3 other traits.
- Doing something that touch lot of people helps.
- Presenting what problem it solves, and what the solution is, in extremly glansable terms is alpha omega. Attention is fickle.
6. Closing/Reflection
- So with just 24h effort xmas hackthon. We reached #1 on PH with a landslide beating YC companies (5m$ funding?). So idea is everything in the AI age (what to build). Execution alone doesnt cut it, altho must be done to the highst standard as well.
- The day after we started to plan a more ambitious sequel. As we just had prooved PMF.
- 2-3weeks later we dropped the sequel and that is now omni. And we already talked to some investors and things look promesing so lets see.
- Final thought: ask questions. how can I ... make a 1m$ dollar....come up with an original tech idea...etc. And keep asking. Push back on obviouse bullshit. Probe probe probe.