- Startup Spells 🪄
- Posts
- 🪄 Nathan Latka's SaaS Data Collection Strategy
🪄 Nathan Latka's SaaS Data Collection Strategy
PLUS: Real Product Hunt Hacks that work even in 2024
Hey,
The Spells Master is back!
Welcome to the 105th issue.
Today's topics:
Nathan Latka's SaaS Data Collection Strategy
A tool to Create 1,000 split screened videos in a fraction of the time
One recommended video on Real Product Hunt Hacks that work even in 2024
Nathan Latka's SaaS Data Collection Strategy
Nathan Latka is a mad man when it comes to being ruthless as an interviewer.
He is very aggressive with his approach in asking the real sauce about Revenue, valuation, and customer data on SaaS companies built from founder audio clips.
He collected all his insights on video on his YouTube Channel so you can see the aggressiveness in action.
He's so good at getting actual data that some founders hate that about him (in private groups) but the insights are mind-blowing indeed.
He used the data from the videos and sells it as a subscription on GetLatka that contains details about revenue from SaaS Founders themselves.
Get Latka SaaS Database
It is costly but you can easily find deals on GetLatka that will make you anywhere from $100k to $1m as an Entrepreneur or Investor.
Get Latka SaaS Pricing
For example, you can reverse-engineering successful startups and build a clone of it is a phenomenal strategy. A few people have made billions using this strategy.
But the funny thing is he spent years building the SaaS Database from 3,156 manual CEO Interviews.
And now you can do it yourself for <$1000-$2000 using AI.
this is awesome. These guys used AI to create a full search engine of My First Million
you can search for any word or phrase, and it'll pull up the exact moment of the podcast where we talked about that thing
— Shaan Puri (@ShaanVP)
9:22 PM • Nov 1, 2022
The process is simple:
Transcribe a podcast using OpenAI's Whisper
Ask AI questions using Custom Prompts
Extract useful metrics like Revenue, Number Of Employees, Valuation from it
Sell it as a subscription product to VC Firms or Entrepreneurs
SaaS like GetLatka are called Database as a Product. They sell like crazy.
You can use build data products for other podcasts that people will pay for.
For example, selling Andrew Huberman's Best Insights from his YouTube Channel as a SaaS using AI. It will definitely cost more as Huberman's podcasts are 3 hours long while Nathan's are 20-30 minutes at most but its a money-printer if you know how to market.
Transcribing Pieter Levels Podcasts
The longer the podcast, the more it costs but the costs are dropping significantly and they will continue to drop.
Definitely read his book "How To Be A Capitalist Without Any Capital" as it has unique insights you won't find elsewhere. He dives deep into how he records a week's worth of podcasts in a day and how he has them scheduled 4 years into the future plus tips on hiring overseas so his podcast gets transcribed for <$1 even before AI came onto the scene.
A tool to Create 1,000 split screened videos in a fraction of the time
Viral Builder AI is a tool that includes a feature of TikTok that's blowing up currently.
Split Screened videos are getting a lot of engagement and this tool helps you create it fast.
One recommended video on Real Product Hunt Hacks that work even in 2024
This video basically covers how to rank on Product Hunt.
Product Hunt (PH) for most people without an audience or community is a shit show.
Its turned into pay to play and the only way to win is having a large audience or pay a hefty $6000 to rank on PH.
This is the only video you need that covers how you can use PH to rank #1.
Top Tweets of the day
1/
The most under discussed immigration policy:
A lesson from Lee Kuan Yew
— George Mack (@george__mack)
7:00 PM • Jul 17, 2024
Onboarding is the place where you must WOW the user in the shortest time possible.
"The key to all consumer is to understand a human need well and then reduce the number of steps it takes to deliver the dopamine."
Singapore knows this.
2/
Exa 竟然还能语义搜索 Twitter/X 的内容,吊炸天 💥,解决了长期困扰我的垃圾 Twitter 搜索问题,感觉利好使用 Twitter 数据做上层产品或者 AI 模型公司:
参见链接:exa.ai/search?q=Perpl…,对应下图搜索:👇
— Tom Huang (@tuturetom)
2:35 PM • Jul 17, 2024
If you want to search for stuff, use this search engine.
You can use Google's Translation Feature to see what it says:
> Exa can even semantically search Twitter/X content, which is amazing 💥 . It solves the spam Twitter search problem that has been bothering me for a long time. I think it will be beneficial for companies that use Twitter data to make upper-level products or AI models: See link: https://exa.ai/search , corresponding to the following search: 👇
Rabbit Holes
Measuring What Really Works on Twitter: Post Timing and Headlines - Phenomenal use-case on how to manipulate any algorithm by an OG so you can essentially reverse-engineer what blows up and what doesn't.
How the 𝕏/Twitter Algorithm Works - Good breakdown of X Algorithm or basically, most text-based algorithms. There are minute differences between different algorithms but they basically just mimic human behaviour. If you like startup content, you will be shown more startup content. Most valuable stuff on startup content will be either shared or linked back to by other people or bookmarked/saved by others. If you are on any social media, don't forget to read it.
LLMs for Dummies - Once you understand the basics of LLMs, you can prompt it better to get what you want.
Until next time,
Your Spells Master!
If you enjoyed this, please consider sharing it with a friend. If a friend sent you this, get the next newsletter by signing up here.
Read all the old case studies here.
Reply