Idea Graveyard

One of my goals is to start a company, and while doing so there are 3 things I’m really prioritizing:
  1. Who I work with.
  2. What I work on. 
  3. How hard I work. 

Considering that a startup is a 6-10 year investment, and ends up being something that consumes your life. I want to be really intentional on these 3 priorities.

Who I work with:

  • I want to make sure I work with the smartest people I can find, who I also consider great friends. I want to be excited to come into the office on Saturday to spend time with them.

What I work on:

  • I want to work on an idea I find exciting, I can develop an emotional connection to, and is challenging. Startups are hard and things go to shit, if you don’t believe in a vision you’ll get tired working on an idea.

How hard I work:

  • Understanding when I’m optimizing for learning, and when I’m optimizing for execution. This will dictate how hard I work and how much time is going to being proactive in learning vs. just dealing with hard tasks that need to be done. Spending time learning how to code something vs. grinding to push out a new feature are two different things.
In September 2019 I started exploring industries and started trying to validate ideas. Here's an Idea Graveyard and an active list of my attempts and failures. 

During the summer of 2019, while I did an internship:
  • I worked on a ton of small projects with the goal of learning to code and going deeper on technical skills. A few projects I started building for fun:
    • A feedback tool for startups to use in teams, it delivers anonymous feedback for weekly performance and how you stacked up against other people on your team.
    • A cryptocurrency market simulator.
    • Lots of small gimmicky tools that used ML.
Once September came I really started going deep into ideas with my friends as co-founders, which led to working on some bigger ideas:

A Remote BDR Monitoring Workspace
  • During my time working and growing an events agency, I had some experience with Business Development Reps and the tasks they do. I also knew a ton of people that employed remote BDRs for their sales teams. Managing a remote BDR had some issues like keeping track of their output and how much time they were actually spending on work, so I explored to see if we could develop a platform that acted as a workspace.
  • Why it failed or I stopped working on it:
    • We started building a prototype, but didn’t want to finish an entire build before we could get a paying customer. 
    • Founders & CEOs didn’t want the tool bad enough that they would pay for it.
    • There were already existing tools in the market that people were really happy with like Hubstaff.
Threads x Tandem
  • Realized our tool was a remote work tool, so we decide to pivot and look at other problems and opportunities in the remote work space we could solve. One of the people we talked to was Andreas Klinger, he's the Head of Remote at Angel List and gave us the idea of combining Threads and Tandem.
  • Why it failed or I stopped working on it:
    • No Founder-Market Fit, we didn’t particularly understand the issues of remote work and didn’t personally face a problem.
    • We weren’t as excited by the idea so we moved on.

  • We had an idea at 2 AM at a Hackathon of using a webcam to track someone's posture and give recommendations on how to improve it. One of the features included, that if your posture was incorrect the screen would blur. We originally started building the idea for fun, but after talking to some people who used Better Back to improve posture we thought it would be an interesting problem to solve
  • Why it failed or I stopped working on it:
    • The technology wasn’t good enough to create a product that users would love.
    • Using the webcam is limiting and can’t track posture as effectively as a wearable can.

Social Podcasting App
  • I wanted to build a social media network based on sharing knowledge learnt from podcasts.
  • Why it failed or I stopped working on it:
    • I learned that the TAM for podcasts was too small, I didn’t want to pursue an idea with that small of a market size.

  • We all hated online courses because they’re too cookie-cutter, and don’t optimize for different types of learning styles. We wanted to build a learning engine that would figure out how a user learned best. Based on that information, when a user would search for an answer, the search engine would throw back a customized list of high-quality content that could be used to learn.
  • Why it failed or I stopped working on it:
    • One of my co-founders got a really crazy job offer in Silicon Valley. It was the founder of a billion-dollar company, starting his new venture working on something incredibly challenging. My co-founder was given an offer as one of the first engineers, and wanted to at least test out a contract. So we decided to break up the team and work on different ideas.

Solving Latency for data transfer in autonomous vehicles
  • One of my friends got back from a retreat in San Francisco, and we were hanging out when he got back. One of the people he talked to in SF was Kyle Vogt (the CTO of Cruise Automation), who told him one of the biggest issues with AVs was latency from gathering data, analyzing it, and sending it to the cloud. We went super deep researching the issue and talking to people, and eventually learned that it was just a first-order problem. The bigger problem was just addressing issues with cloud latency in general.
  • Why it failed or I stopped working on it:
    • We pivoted to solve the larger problem in cloud latency, cloud computing being expensive, and leveraging all the idle computing power available that is unused.
    • We started exploring concepts like edge computing as well.

Using idle devices to crowdsource computing power by turning any device into a server
  • In order to make progress and start executing on this problem, we started off by finding a use case that had a high amount of computation. The goal was to create a marketplace with people who had gaming computers with high specs who would sell idle compute. 
  • We decided to start off with either Labs or Animation Studios, after talking to people in the industry we realized that the people who needed this much compute already had more than abundant access to it with their own local supercomputing clusters. Labs also had very specific security and privacy measures, we didn’t want to get involved with the headache of dealing with privacy in healthcare. While waiting for researchers to respond we decided to put this idea on the back burner, and pursue another problem we found around cloud costs optimization.
  • Why it failed or I stopped working on it:
    • Researchers and Labs had very complex privacy requirements.
    • Did a pivot into another cloud idea.

  • The problem we found was around “cloud optimization.” In short, companies are wasting billions of dollars on extra cloud costs that they don’t need to spend money on, this is due to the wrong settings or configurations in place. We talked to a bunch of companies and realized that cloud spend wasn’t a huge priority, and if it was they already worked to reduce it with a big internal project. We also realized this idea was a one time fix, in order to make it reoccurring we would need to create other value add services and would be trying to solve multiple problems instead of just one big one. While researching the cloud problem, we discovered another problem around the lack of horizontal cloud security monitoring tools that worked for multi-cloud setups.
  • Why it failed or I stopped working on it:
    • The idea was not a scaleable tech company, it would be a one time fix or consultancy. 
    • So we pivoted into Cloud Security 

Creating a multi-cloud security monitoring tool
  • There were lots of small problems that we could solve, including IT managers not being able to view logs of all different clouds in one dashboard or security tool. However, we realized the problem was too small, and we would just be building features. Cloud security is also a really tough space and relies heavily on sales teams, we also had no competitive advantage against industry veterans.
  • Why it failed or I stopped working on it:
    • We had no competitive advantage or unique insights so we decided to work on something else.

Using compounding pharmacies to make cheaper generic drugs
  • People who receive prescriptions for a really expensive drug, can in some cases use a compounding pharmacy. These drugs are crazy expensive, but there are a few pharmacies who often do this as a non-profit and can compound the drug to give it to a patient for a cheaper price. We explored to see if there was a way to add value but still generate a profit.
  • Why it failed or I stopped working on it:
    • The market of people who could use compounding pharmacies based on their prescription is insanely small (TAM under a billion)

  • Learned about Robotic Process automation and we all found it super interesting. Currently working on this idea, possibly reviving Microsoft Clippy.