The CVPR2010 presentation 'Funding: who and how to ask for money' is a rare educational artifact that breaks down the mechanics of early-stage venture capital specifically for the computer vision sector. Rather than a single company pitch, it functions as a strategic manual, using real-world data from two startups: Polar Rose and Kooaba. The deck outlines the startup financing cycle, compares angel versus VC dynamics, and provides specific financial milestones, such as Polar Rose's $5.4 million Series A in 2006. It emphasizes the unique challenges of vision-based startups, particularly the 't…
Key takeaways
- Polar Rose raised $300k in grants and seed funding between 2004 and 2006 before securing a $5.4M Series A (Slide 21).
- Kooaba utilized a mix of PhD-related work, CHF 300k in grants, and CHF 700k in convertible loans for its seed stage (Slide 22).
- The deck warns that computer vision startups often face the 'here is the solution, where is the problem' critique due to being technology-heavy (Slide 19).
- Venture Capital firms are described as having fixed-life funds of 8-10 years, primarily sourced from institutional investors like pensions and foundations (Slide 9).
- A typical angel investment range is cited as $25k-$50k, often requiring networks to make passive investments possible (Slide 8).
- The presentation explicitly advises against NDAs during the initial pitching phase, noting that detailed plans come later (Slide 12).
- Founders are encouraged to know their BATNA (Best Alternative To a Negotiated Agreement) to secure better terms through multiple simultaneous offers (Slide 16).
- The 'VC business model' is defined by the 2/20 rule: 2% management fees and 20% carried interest on profits (Slide 11).
Introduction to the CVPR2010 Funding Guide
The presentation titled 'Learnings from founding a computer vision startup: Chapter 5: Funding' is a historical and strategic document presented at the 2010 Conference on Computer Vision and Pattern Recognition (CVPR). Unlike a standard pitch deck designed to solicit capital for a single entity, this deck functions as a meta-analysis of the fundraising process for deep-tech founders. It utilizes the real-world experiences of two startups, Polar Rose and Kooaba, to illustrate the transition from academic research to venture-backed enterprise.
Slide 1-2: The Funding Framework
The deck opens with a high-level overview of the funding landscape. Slide 2 sets the agenda, covering funding alternatives, the distinction between Angels and VCs, pitching strategies, and the critical question of 'how much' to raise. The inclusion of a 'Barber Shop: We Buy Gold' sign serves as a visual metaphor for the varied and sometimes opportunistic nature of seeking capital.
Slide 3-5: The Startup Financing Cycle and Initial Steps
Slide 3 presents a standard Startup Financing Cycle graph (sourced from Wikipedia). It maps revenue against time, highlighting the 'Valley of Death'—the period before break-even where seed capital and angel investments are vital. Slide 4 introduces 'FFF' (Friends, Family & Fools) and bootstrapping. It notes that a small fresh team can go far with minimal investment, citing an example of an early-stage web product requiring less than $10k per year for computing and tools. Slide 5 acts as a 'Sanity Check,' asking founders if they truly need external funding or if they can continue bootstrapping through consultancy work.
Slide 6-7: External Funding and Grants
Slide 7 focuses on non-dilutive capital, such as soft loans, regional/national grants, and EU research funding (specifically mentioning FP7). A notable inclusion is a warning against business plan competitions, featuring a quote from Steve Blank: 'No one wins in business plan competitions.' This suggests that for technical founders, the opportunity cost of these competitions often outweighs the prize money.
Slide 8-11: Understanding Angels and VCs
Slide 8 defines a business angel as someone with 'experience, energy and cash' who invests their own money, typically in the $25k-$50k range. It also mentions angel networks, which allow for co-investment and passive participation. Slide 9 shifts to Venture Capital, identifying Sand Hill Road as the symbolic hub. It explains that VCs manage fixed-life funds (8-10 years) with capital from institutional investors like pensions and foundations. Slide 11 breaks down the VC business model, specifically the '2/20' structure: 2% management fees for partners and 20% carried interest on profits. It emphasizes that the end goal for a VC is always an exit, whether through a trade sale or an IPO.
Slide 12-14: The Art of the Pitch
Slide 12 provides a pitching checklist: an elevator pitch, a slide deck (which doubles as a business plan), and an optional executive summary. Crucially, it advises founders to 'forget about NDAs' and focus on storytelling, problem-solving, and market size, stating that technology is secondary to the business case. Slide 13 offers specific tips, such as finding a 'champion' inside a VC firm and ensuring materials are sent as PDFs with the recipient's name and date. Slide 14 references a TED talk by David S. Rose as a primary resource for mastering the VC pitch.
Slide 15-17: Negotiation and Valuation
Slide 15 compares Angels and VCs across four dimensions: money, motivation, speed, and control. It warns that VCs 'irreversibly set company on path to EXIT.' Slide 16 introduces the concept of BATNA (Best Alternative To a Negotiated Agreement), advising founders that multiple simultaneous offers are the only way to secure good terms. It bluntly states, 'There is no formula for valuation,' noting it is decided based on 'comfort levels' and that terms are often more important than the valuation figure itself. Slide 17 discusses the quantity of capital to raise, suggesting that while raising 'as little as possible' sounds sound in principle, it risks keeping the founder in 'constant fundraising mode,' which reduces velocity.
Slide 18-19: The Specifics of Computer Vision
Slide 19 addresses the unique aspects of the 'Vision' sector. It notes that research grants are a 'serious option' and that 'visual demos can help pitching.' However, it warns of the 'technology-heavy' trap where founders have a solution but no clear problem to solve. It also observes that in 2010, there was traction in consumer markets but few 'strong success (exit) cases' in computer vision.
Slide 20-22: Case Studies - Polar Rose and Kooaba
The deck concludes with two concrete examples. Slide 21 details Polar Rose's journey: $300k in grants and seed (2004-2006) followed by a $5.4M Series A from Nordic Venture Partners in July 2006, with a follow-up in 2009-2010. Slide 22 details Kooaba's path: CHF 100k from PhD-related work (2006-2007), a CHF 700k seed round via convertible loans and CHF 300k in grants (2008), and a CHF 1.5M total by 2009 through a pre-Series A and further grants.
Slide 23: Resources
The final slide lists influential Twitter accounts and websites for further learning, including VentureHacks (Mark Suster), Fred Destin (Atlas Ventures), Y Combinator, and Seedcamp. It also links to Guy Kawasaki’s 'Art of Bootstrapping.'
What Works in This Deck
The primary strength of this deck is its transparency regarding financial figures. By providing the exact funding trajectories of Polar Rose and Kooaba, the presenters move beyond abstract advice into empirical evidence. The breakdown of the '2/20' VC model and the explicit warning against NDAs are practical pieces of advice that remain relevant today. Furthermore, the distinction between 'technology-heavy' vision startups and market-driven startups is a sophisticated observation that helps technical founders adjust their messaging.
What Is Missing
As an educational presentation rather than a company pitch, the deck lacks a specific 'Ask' or a unified team slide. It does not go into detail regarding the specific computer vision technologies used by Polar Rose or Kooaba, focusing instead on the financial mechanics. There is also no mention of unit economics or customer acquisition costs (CAC), which, while perhaps less common in 2010 for early-stage deep tech, are now standard requirements in modern fundraising decks.
What a Founder Should Copy
Founders should emulate the 'Negotiation' slide's focus on BATNA. The strategy of telling a 'good story to several investors at the same time' is the most effective way to create leverage. Additionally, the 'Pitching Tips' regarding finding an internal champion and being careful with revenue projections (skipping them unless high-quality data exists) are excellent rules for maintaining credibility. Finally, the use of visual demos to bridge the gap between complex technology and investor understanding is a tactic every computer vision founder should employ.
Frequently asked questions
- What were the specific funding amounts for Polar Rose?
- According to slide 21, Polar Rose raised $300,000 through grants, friends, and seed capital between 2004 and 2006. This was followed by a significant $5.4 million Series A round in July 2006 led by Nordic Venture Partners. The slide also notes a smaller follow-up investment from the same firm between 2009 and 2010.
- How did Kooaba finance its early operations?
- Kooaba's strategy was highly diversified. Slide 22 shows they earned approximately CHF 100k from work alongside PhD studies (2006-2007). They then secured a seed round consisting of CHF 700k in convertible loans from banks and FFF (Friends, Family, and Fools), plus CHF 300k in grants in 2008. By 2009, they had reached a total of CHF 1.5 million through a pre-Series A round and additional grants.
- What is the deck's stance on Business Plan Competitions?
- The deck takes a skeptical view of business plan competitions. Slide 7 includes a direct warning: 'Beware of business plan competitions!' and quotes Steve Blank stating, 'No one wins in business plan competitions.' This suggests that the time and effort required for these events may not yield the same long-term value as direct investor engagement.
- How does the deck differentiate between Angels and VCs?
- Slide 15 provides a comparison table. Angels are characterized by providing 'less money,' investing 'for fun,' offering 'quick' processes, and seeking 'no control.' In contrast, VCs provide 'more money,' operate 'professionally,' have a 'slow' process, and demand 'lots of control.' The deck notes that VCs irreversibly set a company on a path toward an exit.
- What specific advice is given for pitching computer vision technology?
- Slide 19 highlights that 'visual demos can help pitching' significantly in this sector. However, it warns that because vision is technology-heavy, founders must be prepared to answer the 'where is the problem' question. It also notes that while there was traction in consumer markets at the time, there were few 'strong success (exit) cases' in the field.
