📚 The Concept of Open Innovation: Inbound and Outbound
In 2003, Henry Chesbrough published Open Innovation: The New Imperative for Creating and Profiting from Technology and destroyed a 50-year-old dogma: that true innovation only happened inside proprietary labs. The argument was straightforward — useful knowledge is distributed around the world, and no company, no matter how large, can monopolize it. Trying to do everything in-house is expensive and increasingly thankless.
→ Inbound Open Innovation
The company absorbs external knowledge to accelerate its own development. Licensed technologies, external talent, startup partnerships, open source, published challenges. The logic: if someone has already solved the problem, why reinvent it?
Example: Procter & Gamble with the Connect + Develop program—50% of innovations came from outside the company (Lafley’s goal, 2000s).
← Outbound Open Innovation
The company monetizes internal knowledge that it doesn’t use in its core business. Licensed patents, spin-offs, joint ventures, selective research publication. The logic: IP sitting in a vault creates no value—it’s better to give it to someone who will use it.
Example: IBM licenses more than 3,000 patents a year, generating billions in royalties from technologies that aren't core to its main business.
📊 Data and Research
- • 78% of Fortune 500 companies have some open innovation initiative (Deloitte, 2023)
- • Companies with structured OI programs launch new products 2.4x faster than closed ones (MIT Sloan, 2022)
- • The average cost of internal R&D is 3-5x greater than external partnerships for the same output (BCG Innovation Survey)
- • R&D intensive in the closed model (e.g., Xerox PARC) generated technologies that others monetized — Ethernet, GUI, laser printing
📈 Corporate Venturing and CVC
Corporate Venture Capital (CVC) is the most sophisticated instrument for inbound open innovation. The company invests capital in external startups to obtain financial return AND strategic access to technologies, markets, and talent. It’s not philanthropy or sponsorship—it’s a position of future optionality. Modern CVC has evolved from simply "let’s invest in startups" into an integrated competitive intelligence system.
🕐 Evolution of Corporate Venturing
Passive CVC (1990s)
Minority investment in startups, without integration with the business. The main goal: financial returns. Typical result: the startup grows, but the large company learns nothing. Model discontinued in most cases after the dot-com crash.
Strategic CVC (2010s)
Investment tied to strategic theses: "I want to learn about X technology" or "I want access to market Y." Intel Capital, Google Ventures, Salesforce Ventures. The investment is conditional on commercial or technological collaboration. The startup gets money + a customer/distribution. The corporation gets intelligence + an acquisition option.
Integrated Ecosystem (current)
CVC as one layer of a broader system: in-house accelerator + CVC + M&A program + innovation hub. Natura &Co, Itau BBA, Embraer X. The startup is more than an investment — it’s an external lab. If it proves itself, acquire it. If not, you learned cheaply.
✓ CVC That Works
- ✓Clear thesis: which technologies/markets we want to understand
- ✓Dedicated team with autonomy to make decisions (doesn’t need 5 approvals)
- ✓Fast response SLA for startups (max 4 weeks to a decision)
- ✓Mechanism for transferring learning to the core business (meetings, workshops)
- ✓Predefined exit criteria (acquisition, divestment)
✗ CVC That Fails
- ✗Created to “look innovative” to the press and competitors
- ✗CVC team without access to the C-level to close commercial partnerships
- ✗Invested startups can’t sell to the very companies that invest in them
- ✗Conflict of interest: startup reveals IP and loses its competitive advantage
- ✗Wrong metric: evaluating CVC with the same ROI as traditional private equity
👥 Partnerships with Startups and Scaleups
Partnering with startups isn't about being "cool" or appearing in the innovation rankings. It's about access speed, technology, and business models that a large company can’t create internally in the time required. The challenge isn’t finding startups — it’s structuring the relationship so it creates real value for both sides without killing the startup in the process.
🕐 The 4 Partnership Models
1. Customer/Supplier
The company buys the startup’s product/service. The simplest model. The startup gets revenue and a reference. The company gets a quick solution. Risk: dependence on a fragile supplier.
2. Co-development
Jointly develop a solution. The company brings a real problem + data + distribution. The startup brings technology + speed. IP split agreed in advance. More complex, more value.
3. Channel/Distribution
The company distributes the startup’s product to its customer base. Or the startup distributes it to a market the company can’t reach. Synergy between the installed base and agility.
4. Accelerator Program
The company sets up an accelerator (e.g., Wayra by Telefónica, Liga Ventures). Startups receive mentoring, resources, and access to customers. The company gets a pipeline of technologies and an option to invest in or acquire them.
💡 The Pilot Agreement: how to structure it
The biggest bottleneck in the partnership is the eternal pilot—the startup stays “in pilot” for 12 months with no purchase, no expansion, and no formal decision to stop. This kills the startup and the relationship. The solution is the Structured Pilot Agreement:
- • Fixed duration: 60-90 days, never open-ended
- • Predefined success metric: "if X happens, we buy"
- • Internal sponsor with purchasing decision-making authority (not just for a pilot)
- • Continuation budget already approved in principle
- • Clear exit: if it doesn't work, the company lets the startup move forward
🇧🇷 Brazilian Case Studies
🏫 Universities, R&D, and Grant Programs
University research is Brazil's most underused asset in open innovation. While the triple helix (government + business + university), described by Henry Etzkowitz and Loet Leydesdorff, works well in the U.S., Germany, and Israel. In Brazil, the gap is structural: researchers don't know how to reach companies, and companies don't know how to absorb basic research. But those who learn to navigate this ecosystem gain a lasting competitive advantage.
📚 University-Business Collaboration Mechanisms
Direct R&D contracts
A company funds specific research at a university lab. It defines the problem; the university chooses the method. Access to results through a contract. Works best for well-defined technical problems (materials, processes, pharmaceuticals). Typical timeline: 12-36 months.
FAPESP, CNPq, EMBRAPII funding calls
The government subsidizes part of the cost of collaborative R&D. FAPESP PITE (Partnership for Technological Innovation): the company contributes 50%, and FAPESP covers 50%. EMBRAPII: a subsidy of up to 1/3 of the project. Result: the company gets access to cutting-edge university research at a 50% discount, with IP protected.
Corporate residency programs
Researchers spend 6-12 months embedded in the company. The company learns academic language, and the researcher learns business language. Examples include the MIT Industrial Liaison Program and the Stanford Center for Professional Development. In Brazil, some companies like Petrobras and Embrapa have similar programs.
💡 How to get started without bureaucracy
Don't start with grant applications — start with people. The fastest path:
- 1. Identify 2-3 reference labs in your sector (USP, UNICAMP, UFMG, POLI). Use Google Scholar to find the researchers most cited on your problem.
- 2. Get in direct contact with the researcher (not the university administration). Invite them for a visit to your company.
- 3. Start with a small project: a capstone or sponsored dissertation (R$ 20-50k). Learn the process without risk.
- 4. If it works, scale up to FAPESP PITE or a larger R&D contract.
🌐 Crowdsourcing, Platforms, and Open Challenges
Crowdsourcing innovation is the act of publish a problem for the world to solve. The concept was popularized by Jeff Howe in Wired Magazine in 2006 and evolved from simple design contests into sophisticated collaborative problem-solving platforms. Chesbrough's premise applies here too: if knowledge is distributed, the solver of your toughest problem could be in Singapore, the countryside of Maranhão, or an engineering student's garage.
🏆 Reward Platforms
InnoCentive, HeroX, Kaggle. The company publishes a technical challenge with a cash prize. Solvers from around the world compete. The company pays only for a solution that works. Typical ROI: 10-50x vs. internal R&D (MIT, 2019).
Example: NASA used InnoCentive to solve a solar radiation prediction problem. It was solved by a retired physician with no background in astrophysics.
📈 Corporate Hackathons
48-72 hours, a real problem, a prototype as the deliverable. Works for digital product innovation, internal processes, and new business models. The classic mistake: a polished hackathon with no results implemented. The solution: pre-commit an implementation budget for the top 3 finishers.
Example: Santander Hackathon Brasil — the winning solution became a real product (Santander Empresas digital) in 8 months.
🌞 Strategic Open Source
Release internal technology as open source to attract external contributions and create a market standard. Google with Android and TensorFlow, Netflix with Chaos Monkey, Meta with React. The company gives up exclusivity but gains an ecosystem.
Logic: If many players adopt your technology, you set the standard. Whoever sets the standard doesn’t pay royalties — they receive them.
📋 When to Use Each Mode
| Format | Ideal problem | Cost |
|---|---|---|
| Prize platform | Technical, definable, clear metric | Medium (prize) |
| Hackathon | Digital, product, process | Low |
| Open source | Infrastructure, platform, standard | High (initial) |
| Co-creation with users | Experience, consumer product | Low |
⚠️ Pitfalls: The "Not Invented Here" Syndrome and Intellectual Property
Open innovation fails silently. There is no dramatic moment of collapse — it simply doesn't happen, because the organization has antibodies against what comes from outside. The most famous of these antibodies is the "Not Invented Here" syndrome (NIH — Not Invented Here), first documented by Katz and Allen in MIT Sloan Management Review in 1982. The other is naive intellectual property management, which creates more litigation than value.
🚫 NIH Syndrome: How to Identify and Address It
NIH is not conscious resistance. It is a cognitive bias: people overestimate the value of what they created and underestimate what came from outside. R&D teams with a long history of success are the most affected. Typical manifestations:
Warning Signs
- ✗"This is very simple; we can do it better in-house"
- ✗"The startup doesn't understand the specifics of our industry"
- ✗Stricter evaluation criteria for external solutions than for internal ones
- ✗R&D teams blocking the purchase of solutions that “compete” with their roadmap
How to overcome
- ✓Recognize NIH as a business risk (not a cultural weakness)
- ✓Include "evaluated external solutions" as an R&D team metric
- ✓Create “technology scout” roles with incentives to bring in the best from outside
- ✓Celebrate successful external adoption as much as internal innovation
⚖ Intellectual Property in Open Innovation
The biggest fear companies have about open innovation is losing IP. The biggest fear startups have is having their IP stolen. Both fears are legitimate and often poorly addressed. Chesbrough argues that IP should be viewed as asset that needs to circulate, not as a fortress to defend. Idle IP generates no returns.
💡 Chesbrough's Lesson
Chesbrough’s most provocative conclusion in Open Business Models (2006): "The company with the best business model wins, not the company with the best technology." IP is not a lasting competitive advantage. What lasts is the ability to orchestrate the ecosystem—partners, startups, universities, platforms—to continuously create and capture value. Companies that overprotect end up protecting assets the market has already outgrown.
📋 Module Summary
Next Module:
3.4 - Organizational Ambidexterity