4. Learnings from founding a Computer Vision Startup
FFF (friends, family & fools)
A small investment can take a small fresh team far
E.g. early stage Web product <$10k/year for computing and tools
Bootstrapping:
Consulting work on the side (that hopefully also builds your product)
Hire young, cheap, and hungry
Minimum viable product (don’t build more than an absolute minimum)
5. Learnings from founding a Computer Vision Startup
Sanity check!
Do you really need to raise external funding?
can you bootstrap of own investment, consultancy work, part time...
What is the money for?
after “how much?”, this is the next question investors will ask
7. Learnings from founding a Computer Vision Startup
Grants & Research Funding
Soft loans
Regional & National grants
EU research funding (FP7 etc)
Beware of business plan competitions!
“No one wins in business plan competitions“
-Steve Blank
Flickr:denniswong
8. Learnings from founding a Computer Vision Startup
Angels
What is a business angel?
someone with experience, energy and cash
invests own money
typical range 25-50k
Angel networks
groups of angels that co-invest
needed since each angel has limited funds/investment
makes passive investments possible
Flickr: onkel_wart
9. Learnings from founding a Computer Vision Startup
Venture Capital
Fixed life funds (~8-10 years)
Money comes from institutional
investors
pensions, universities, foundations, individuals
Types
regional vs. (inter)national, early vs. late,
branch specific
Flickr: markcoggins
11. Learnings from founding a Computer Vision Startup
Organization
partners & associates
End-goal is an EXIT
Trade sale
IPO The VC business model (2/20)
Expectations management fees - partners get ~2% of invested capital
carried interest - partners’ management company get ~20% of profits
some form of control
most investments fail
the few that succeed should give high multiples
>>> High risk is a part if the game but risk level depends on timing, spread of
portfolio and financial climate.
Flickr: sophistechate
12. Learnings from founding a Computer Vision Startup
Pitching investors
Checklist:
Storytelling + elevator pitch
+ slide deck (doubles as
presentation and business plan)
Problem & solution + exec summary (optional, very
short!)
(technology is secondary)
- forget about NDAs
- detailed plan comes later
Show a market
Flickr: akrobat77
13. Learnings from founding a Computer Vision Startup
Pitching tips
Pitching angels
Pitching VCs
chemistry: it’s about relationship and trust
prepare: look at their portfolio & profiles
often a group is needed (don’t worry, they
find a “champion” on the inside bring in “friends”)
partners decide together - convince them all! if decisions don’t come quickly, it is a “no”, -
pitch at partner meeting will be short, be move on.
flexible
Be careful with revenue projections! (skip unless really good data exists)
If you send material or a deck, make it .pdf with recipients name and date.
If multiple investors, work the “lead” investor into agreement, then offer others same terms.
14. Learnings from founding a Computer Vision Startup
David S. Rose on pitching to VCs
http://www.ted.com/talks/david_s_rose_on_pitching_to_vcs.html
15. Learnings from founding a Computer Vision Startup
Angels or VC?
if possible: look for integrity,
Angels VC
“you can’t fire your investors”
less money more money
quick exit vs. big exit
for fun professionally
quick slow beware that VC irreversibly sets
company on path to EXIT
no control lots of control
seek investors who agree with your definition of success
16. Learnings from founding a Computer Vision Startup
Negotiation
Know your BATNA*
“to get good terms you need multiple offers, to get multiple offers
you need to tell a good story to several investors at the same time”
There is no formula for valuation, it is
decided on comfort levels. Actual terms are
more important than valuation and dilution.
* Best Alternative To a Negotiated Agreement
17. Learnings from founding a Computer Vision Startup
How much to raise?
“It depends”, 3 different views:
1) As little as possible
2) As much as you can (against reasonable equity, of course)
3) Whatever you can get and get on with building stuff
1) Sound in principle but risks spending time in “constant fundraising mode”
which takes lots of energy from building stuff
2) “you are always going to need more than you think”
3) Fundraising is a distraction, time spent is distraction and reduces velocity
Tip: If possible raise all at once, avoid tranche investments
19. Learnings from founding a Computer Vision Startup
What’s special about Vision?
Research grants can be a serious option
Visual demos can help pitching
Currently lots of traction in consumer markets (people and businesses
asking for vision) but few strong success (exit) cases.
Vision is technology heavy - you will encounter the “here’s the solution
where’s the problem” comments
21. Learnings from founding a Computer Vision Startup
Polar Rose: How we did it
Grants, Friends & Seed
$300k over 2004-2006
Series A
$5.4M in July 2006 from Nordic Venture Partners
-> smaller follow up in 2009-2010 from Nordic Venture Partners
22. Learnings from founding a Computer Vision Startup
Kooaba: How we did it
Grants, work besides PhD
~ CHF 100k over 2006-2007
Loans seed round (bank + FFF) & Grants
~ CHF 700k seed (convertible loans), CHF 300k Grants (2008)
Pre series A round & Grants
~ CHF 1.5 Million total (2009)
Series A now
23. Learnings from founding a Computer Vision Startup
Resources
Twitter:
@venturehacks (Startup advice on VC and entrepreneurship)
@msuster (Mark Suster, entrepreneur gone VC, also at http://bothsid.es)
@fdestin (Fred Destin, VC at Atlas Ventures)
Seed funding:
Ycombinator (North America) http://ycombinator.com/
Seedcamp (Europe) http://www.seedcamp.com/
Angels:
http://venturehacks.com/angellist (list of angels on Twitter)
Bootstrapping:
http://blog.guykawasaki.com/2006/01/the_art_of_boot.html (Guy Kawasaki on bootstrapping)