ICYMI the Startup markets are getting hotter in the Data Science space. Every time I turn around, some small company got millions of dollars in startup funding. It used to be a company with an algorithm or data science library but now it’s Data Science platforms. These platforms are suddenly all the rage and many new entrants are racing to gain market and mind share.
The above image from FundersandFounders.com really captures a successful startup from inception to IPO. Most interesting for me is how the ownership “pie” is cut over time. If you’re the Founder, you first start out with a 50/50 share with your Co-Founder. Then you get some Seed money, say from an Angel Investor like Howard, which takes a small % of ownership.
As the Startup grows and matures it should attract more VC money and the ownership pie changes. With every VC investor, you sell parts of your company. This is incredibly important if you want to maintain control of your company and should be carefully analyzed.
My personal opinion is that you can do all this without VC money but it will be harder and take a longer time. It could take decades and in this industry time is not your friend. The market is so hot that your competitors will fill your weaknesses in the market within a quarter or shorter. So in essence you really need startup funding from VC’s to be agile and build/keep your market share. Just keep an eye on those term sheets and make sure that the “pie” is big enough for everyone.
I’ve been following Howard and Fred on and off over the years. They’re old Web 2.0 veterans like me and have kept a blog going for over 10 years. It’s always nice to see continuity and recently I found a great interview of Fred at MIT on Howard’s blog. It’s just under an hour and gives advice to students interested in the Venture Capitalist (VC) field.
What caught my ear (and Howard’s too) was how his wife (Gotham Girl) is his biggest cheerleader. What does he exactly mean by that? She was the one that supported and believed in him to go out into Venture Capitalism.
With any big risk you take in life, it makes your life so much easier to have someone behind you, cheering you on. Like Fred, my wife is my biggest cheerleader. She encouraged and supported me to step out of the Engineering world and into the Startup world.
You won’t believe how much that mattered to me then and how much it still matters today.
Fred’s actual talk is about 25 minutes or so, but the Q&A part is gold. I’m going to watch this a few times just to glean as much wisdom as I can!
Other Interesting Bits
- The Founder(s) need to be like salespeople, they have to sell their idea to VC’s, customers, etc
- You need to build up well diversified startup portfolio, some will fail and you’ll lose your seed investment
- It’s a good idea to help negotiate the Series A and even B rounds, because that’ll help you get your exit
- Opportunities abound everywhere
When you’re dealing with a classification problem in machine learning, good labeled data is crucial. The more time you spend labeling training data correctly, the better. This is because your model’s performance and deployment will depend on it. Always remember that garbage in means garbage out.
Thoughts on labeling data
I recently listened to a great O’Reilly podcast on this subject. They interviewed Lukas Biewald, Chief Data Scientist and Founder of CrowdFlower. CrowdFlower provides their clients with top notch labeled training data for various machine learning tasks, and they’re busy!
The few bits that caught my ear were how much of the training data is used in deep learning. They’re also seeing more image labeled data for self driving cars.
The best part of the interview as Lukas’s discussion on using a Raspberry Pi with Tensor Flow! How cool is that?