3. Learnings from founding a Computer Vision Startup
How many?
1-4 “reasonable” size of founder team
1 founder: total control but expensive, lonely and slow
2 founders: the magic number. “one builds, one sells”
3-4 founders: can make a great dev team but leader needed
>4 founders: crazy unless some are “passive”
“Everyone obsesses with dilution from investors.
The biggest dilution comes from co-founders.”
– @msuster
4. Learnings from founding a Computer Vision Startup
2 founder dynamics Jobs and Wozniak
Allen and Gates
Hewlett and Packard
Larry and Sergei
Yang and Filo
“The ideal founding team is two individuals, with a
history of working together, of similar age and financial
standing, with mutual respect. One is good at building
products and the other is good at selling them”
– @venturehacks
http://venturehacks.com/articles/pick-cofounder
Flickr: oskay
5. Learnings from founding a Computer Vision Startup
Finding founders & early employees
Share ideas, be open
Co-founders need to be people you trust, preferably people you
worked with before
You want broad skills and doers
Controversial but interesting:
vesting founder shares (VCs will ask for this)
6. Learnings from founding a Computer Vision Startup
Size matters
Output is not linear in # employees
Most efficient (development) team is ~4
Over ~15 “formal” organization and information channels becomes
increasingly important
Avoid hiring admin staff or positions that don’t “produce” anything
(non-coding proj manager, office manager...) as long as possible.
Ideally, never hire such persons.
7. Learnings from founding a Computer Vision Startup
Flickr: yodelanecdotal
Part 2: Hiring staff
8. Learnings from founding a Computer Vision Startup
Where to find them?
1. By recommendation (both ways!)
2. Events and conferences
3. LinkedIn!
Avoid recruitment agencies at all cost.
Waste of time and money.
Never hire unless you absolutely must.
Are you sure you need to hire now?
Flickr: bouldair
9. Learnings from founding a Computer Vision Startup
Interviewing and testing
Early on, look for people that are bright, broad with potential to grow
Always test developers
Don’t underestimate team fit (personality)
Never hire without trial period
Computer vision team should have a mix of skills
“theoretical”-”practical” and people that can make demos.
10. Learnings from founding a Computer Vision Startup
Stock options and motivation
Motivation
Money is not motivation (competitors ALWAYS pay more)
Good motivators are similar to founders’ motivators (build something, dynamic
organization, make a difference, ...)
Warning: people with “big-co” motivation (titles, salary, career, benefits, “security”)
Stock options
Good idea if exit is the goal
If possible create options pool post investment
Beware: rules and taxes vary a lot between countries!
12. Learnings from founding a Computer Vision Startup
What’s special about Vision?
Academic research groups can be a great “extension” to your R&D
team
Small vision teams can go far, no need to overstaff
14. Learnings from founding a Computer Vision Startup
Polar Rose: How we did it
1. Founder + key early employees
2. Built 2 dev teams (vision + infrastructure)
Hiring process:
Initially friends and connections from university
Networks of a few key employees (especially France and Poland)
LinkedIn - search, search, search
Trial period in ALL contracts
Stock options (legal & tax mess with multi-national team)
15. Learnings from founding a Computer Vision Startup
Kooaba: How we did it
Founders + 1 key early employee (first employee needed to be changed, lost lots of time)
Built two dev teams (vision + interfaces (web, mobile))
Was hard to find initial employees
Work permit problems
Hiring process:
Initially friends and connections from university
Networks of a employees We are hiring later this year!!
LinkedIn - post job offer (299 is cheap)
Sales & Marketing hired from customer (EMI Music)
Employees are involved in interviewing process
Recently: testing day
Stocks (promised, formalized these days)