The Power of Skunkworks Projects (a conversation with ChatGPT)
The Power of Skunkworks Projects (a conversation with ChatGPT)
S2-E7 (2 April 2023)
If you were forwarded this email or if you come from Linkedin, you can sign up to receive an article like this every fortnight.
In this episode of the newsletter, I used ChatGPT as a co-author to cover the topic of innovation and new product development in large organizations.
I chose this topic because I know it very well, so I can check that the text produced by the machine is correct and that there are no hallucinations. ChatGPT does not provide a list of references to check the accuracy of its claims. Therefore, it is prudent to use it as a copilot on topics you master and avoid relying on artificial intelligence as an omniscient and infallible entity.
Large companies must grow
No large company can stand still. Shareholders expect it to grow daily and generate revenue to increase the value of their shares and collect dividends.
There are essentially two ways a company can grow. Expand penetration in markets where it already has a presence or enter new markets.
In either case, a make-or-buy strategy can be embraced. In other words, the company:
can decide to invent a new product from scratch by leveraging its resources,
or it may choose to buy another company to expand its range of offerings and enter new markets.
Large companies are generally very good at creating incremental improvements to existing products, enhancing performance, or creating variants to preside over a specific customer niche. In contrast, they are more unlikely to create something radically new with the same predictability.
Apple launched the iPod in 2001, the iPhone in 2007, and the Watch in 2015, and it continues to announce new models of these products every six months. What will be the next revolutionary product? Maybe there won’t be one. ChatGPT explains why:
Nicola. Why is it so difficult for large companies to create successful new products?
ChatGPT. Creating new successful products can be particularly challenging in large companies for several reasons:
Bureaucracy and hierarchy: In large organizations, there are often multiple layers of management and complex approval processes, which can slow down decision-making and stifle innovation. This may hinder the agility needed for successful product development.
Risk aversion: Larger companies tend to be more risk-averse as they have more to lose and a greater focus on protecting their existing market share. This can result in a reluctance to invest in new, unproven ideas or to disrupt their current business model.
Internal competition for resources: Large companies have numerous departments, projects, and initiatives competing for limited resources, such as funding, personnel, and executive attention. This can make it difficult to secure the necessary support for new product development.
Inertia and resistance to change: Established companies often have entrenched ways of doing things, which can create resistance to new ideas and processes. This can make it difficult to implement the organizational changes needed to support innovation.
Focus on short-term results: Publicly traded companies, in particular, are under pressure to deliver short-term results to shareholders. This can lead to a focus on incremental improvements to existing products and services rather than investing in more transformative, long-term innovations.
Siloed departments: In large organizations, departments often operate in silos with limited communication and collaboration between them. This can hinder the cross-functional teamwork needed to develop innovative products that address customer needs holistically.
Difficulty scaling innovation: While innovation might occur in pockets within a large company, it can be challenging to scale these efforts across the entire organization, which may limit the overall impact of new product initiatives.
Talent retention: Large companies may struggle to attract and retain top talent, who might be more drawn to the agility, autonomy, and innovation potential offered by smaller companies or startups.
Diminished sense of ownership: In a large company, individual employees may feel less connected to the overall success of a new product, leading to a reduced sense of responsibility and commitment to the project.
Despite these challenges, large companies can still create successful new products by adopting strategies that foster innovation, collaboration, and agility, such as establishing skunkworks teams, implementing agile methodologies, and cultivating a culture of continuous improvement.
ChatGPT’s response is - as one might expect - academic, but it provides a practical overview of the topic. Many of the reasons listed can be summed up in two words: corporate culture.
Large organizations attract people who feel comfortable in a predictable and reassuring environment. On the other hand, our educational system is still structured to produce skilled workers destined to perform repetitive tasks. Precisely the opposite of what is required when experimenting with creating something new, and it is, therefore, necessary to come to terms with uncertainty, welcoming it as a traveling companion.
The best strategy for creating new products
Nicola. What is the best strategy to create a new product inside a corporation?
ChatGPT. There is no one-size-fits-all strategy to create a new product within a corporation, as the best approach will depend on various factors such as the company’s size, culture, resources, and industry. However, there are several key principles that can guide the development of a successful product:
Foster a culture of innovation: Encourage a work environment that supports creativity, experimentation, and risk-taking. This helps to create a mindset where employees feel empowered to suggest and explore new ideas.
Clearly define objectives and goals: Before starting product development, establish clear objectives and goals that align with the company’s vision and strategy. This helps to guide the team’s efforts and provides a framework for decision-making.
Assemble a diverse, cross-functional team: Include individuals from various departments, such as engineering, design, marketing, and sales. This ensures that different perspectives and skill sets are represented and encourages collaboration and effective communication.
Adopt an agile product development process: Implement a flexible, iterative approach to product development, allowing for rapid testing, learning, and adaptation. This can include methodologies such as Scrum, Lean, or Design Thinking.
Encourage customer-centricity: Keep the customer at the center of the development process by conducting market research, gathering user feedback, and continuously iterating based on customer needs and preferences.
Allocate resources wisely: Ensure that the project has adequate funding, personnel, and time to succeed. This may involve securing executive sponsorship or creating a skunkworks team to operate outside of the usual organizational structure.
Foster open communication and collaboration: Promote transparency and knowledge-sharing among team members and across the organization. This can involve regular status updates, cross-functional meetings, and open channels for feedback and suggestions.
Track progress and measure success: Establish key performance indicators (KPIs) and monitor them regularly to ensure the project is on track and meeting its objectives. This allows for timely adjustments and improvements as needed.
Learn from failures: Recognize that not every product will be a success, and use failures as learning opportunities to inform future projects and enhance the company’s overall innovation capabilities.
Celebrate successes: Acknowledge and reward the hard work and achievements of the team when a new product is successfully brought to market. This helps to cultivate a positive culture and motivate employees to continue innovating.
Again, the answer can be considered correct. Interestingly, once again, most of the strategies listed require a change in corporate culture. No wonder. The academic and business literature on innovation in recent decades speaks extensively about the impact of organizational dynamics on a company’s ability to remain competitive and create new products and services. If ChatGPT had been trained 50 years ago, it probably would have responded that it was necessary to establish an extensive research and development center along the lines of Bell Labs or Xerox Park.
What is a skunkworks project?
ChatGPT explicitly mentions skunkworks projects in both responses. It uses this expression because I started this conversation about innovation in large companies by asking for descriptions of these types of initiatives.
The machine takes context into account, so the previous answers influence the flow of the conversation. If I had phrased the interview by asking the questions in a different order, I would have gotten similar responses in substance but slightly different in wording.
Nicola. Can you describe a skunkworks project?
ChatGPT. A skunkworks project is an initiative or undertaking that operates outside of the normal organizational structure and is typically characterized by a small, dedicated team working in a highly autonomous manner. The term “skunkworks” originated from Lockheed Martin’s Advanced Development Programs (ADP) division, which was responsible for developing cutting-edge aircraft designs during World War II.
In a skunkworks project, the team is often granted significant freedom to experiment, innovate, and take risks without the usual bureaucratic constraints. This allows for rapid development and progress, often leading to breakthroughs and disruptive innovations.
For example, imagine a large tech company seeking to develop a groundbreaking new wearable device. They might establish a skunkworks project by selecting a small, highly skilled team of engineers, designers, and product managers to work on this initiative. The team would be given a separate workspace, a flexible budget, and the authority to make decisions without seeking approval from the larger organization.
By operating in this way, the skunkworks team could rapidly iterate on their designs, test new ideas, and pivot when needed without being bogged down by the usual corporate processes. The end result might be a revolutionary product that reaches the market more quickly and effectively than if it had been developed through the company’s standard R&D process.
It’s true. Skunkworks projects are an effective strategy because they allow a compromise between the cumbersome organizational dynamics of a large company and the need to have an agile approach when exploring new business ideas.
How to organize a skunkworks project
Nicola. What is the best way to organize a skunkworks project?
ChatGTP. Organizing a successful skunkworks project requires careful planning and execution. Here are some best practices to consider when setting up a skunkworks project:
Clearly define objectives and scope: Establish the goals, objectives, and scope of the project, ensuring they align with the company’s overall vision and strategy. This provides a clear sense of direction and purpose for the team.
Obtain executive sponsorship: Secure support from top-level management to ensure the project has the necessary resources, autonomy, and backing to succeed. This can also help to overcome potential bureaucratic obstacles.
Assemble a dedicated, cross-functional team: Select a small, highly skilled team with diverse expertise, including members from different departments such as engineering, design, marketing, and sales. This encourages collaboration and ensures a well-rounded approach to problem-solving.
Grant autonomy and flexibility: Give the team significant freedom to make decisions, take risks, and iterate quickly without being bogged down by the usual corporate processes. This can foster a more agile and innovative working environment.
Provide a separate workspace: Designate a dedicated physical space for the skunkworks team, separate from the main organization. This can help minimize distractions, encourage creative thinking, and strengthen team cohesion.
Establish a clear timeline and milestones: Set a realistic timeline with specific milestones to track progress and keep the team focused. This allows for more efficient use of resources and helps maintain momentum.
Maintain open communication: Encourage transparent communication within the team, as well as with the broader organization and executive sponsors. This can involve regular status updates, meetings, and feedback loops to ensure everyone is informed and aligned.
Implement an agile development process: Utilize agile methodologies, such as Scrum, Lean, or Design Thinking, to enable rapid iteration, testing, and learning throughout the project. This can help the team adapt quickly to new information or changing circumstances.
Focus on customer-centricity: Ensure the project remains centered on addressing customer needs and preferences by incorporating market research, user feedback, and customer insights throughout the development process.
Measure success and learn from failures: Establish key performance indicators (KPIs) to evaluate the project’s success and learn from any setbacks or failures. This can help inform future skunkworks projects and improve the company’s overall innovation capabilities.
By following these best practices, companies can increase the likelihood of a successful skunkworks project, leading to innovative products and solutions that drive growth and competitiveness.
This response is also agreeable but very general and of little use in creating a plan of action. Let’s try one step further: ChatGPT cannot help this time because it lacks the necessary experience.
If a large company asked me to leverage my many years of experience as a startup founder and advisor to put together a skunkworks project to develop a new software-based product (my area of expertise), I would proceed in four steps:
Step 1. Minimum Viable Product
The goal of the first phase, which could last 3 to 6 months, is to explore a business opportunity. It is then necessary to do user research, prototyping, and testing to arrive at a minimum viable product (MVP) that demonstrates the idea’s market potential.
The startup team should mix the client’s industry know-how sector with skills in product management, design, and go-to-market. Therefore:
I would select a group of 2 or 3 company employees and have them undergo training to give a general smattering of methods such as design thinking, lean startup, product-led growth, etc. I would train the internal team by giving preference to those coming from sales and marketing. These people, moreover, should be formally allocated to the skunkworks project full-time without penalizing their career path.
I would pair the group of employees with an external team (we’ll call it the skunks’ team) of super freelancers consisting of a product manager, a UX/UI designer, a user researcher, a growth hacker, and a developer who is an expert in no-code and low-code tools. The skunks’ team is tasked with taking leadership of the project and operationally carrying out all ideation, design, development, and go-to-market activities. At the same time, it will incorporate company employees and continue their training to make them autonomous.
Phase 2. Kick-off, Roadmap, and Funding
Once the mixed team has defined the product to be developed, it is necessary to determine how to move the project forward. At this stage, it is essential to:
Decide whether it will be a spinoff or a business unit. The difference is crucial. In the first case, the goal is to create a new company, and it is necessary to incorporate the company and recruit the team from scratch. In the second case, the skunkworks project is to incubate a product that will grow within the parent organization: people need to be recruited following the parent company’s procedures, it is helpful to adopt some organizational processes right away, budget constraints, compensation policies, and so on need to be taken into account.
Share vision and roadmap. Developing an MVP, validating it, and creating an investor pitch might be enough to secure a small investment for a startup. In a large company, greater formalization and a few more documents are often needed to share the project with stakeholders and create the necessary alliances. A roadmap with a plan of goals you intend to achieve regarding product development and commercial growth is generally a good way to dispel many doubts. A skunkworks project is a special initiative but should not work in complete isolation with the risk of making enemies.
Secure funding. Who allocates the budget to develop the product? In the case of a startup, actual venture capital funding is needed. If, on the other hand, the skunkworks project will generate a business unit, it is necessary to identify the cost centers and go through all the internal procedures to allocate the budget (in large companies, it can become a way of the cross).
Step 3. Team setup
In this phase, two key activities need to be carried out to ensure the future of the project, namely:
Building the product and growth teams. Since the skunks team is temporary, company employees must take on the team lead role, and new resources must be recruited to form the product and growth marketing teams.
Creating the software development team. If you decided to create a spinoff in the previous step, you must hire a CTO and build the software development team. Otherwise, an in-house tech lead should be identified and a sufficient number of programmers seconded.
Step 4. Product Development and Handover
If the first three phases have been successful, all that remains is to work toward implementing the roadmap and launching the product. This phase, which should last from 9 to 18 months, depending on the complexity of the initiative, requires a lot of discipline and the ability to stay focused on the goal at hand.
This is when project leadership is gradually transferred from the skunks’ team to the team that will take the product forward and grow it either as a startup or as a business unit. It is time for the skunks’ team to let the baby walk independently.
Skunkworks projects and venture building
You are right if you think that skunkworks projects look like the venture-building initiatives that all the big consulting firms have launched in recent years. But there are two significant differences, namely agility, and misalignment of interests.
Large consulting firms are heavyweight firms with an established business model: they sell their employees’ time to clients and replicate projects already done for other clients. Doing the same project repeatedly lowers the risk of failure and increases margins. Unfortunately, this model has little to do with the agility needed when inventing a new product or service from scratch. A startup is in the realm of uncertainty and novelty, not repetition and best practices.
To the total absence of agility, it must be added that a consulting firm’s reward system is margin over man-days, so a venture-building project must be massive and long and occupy many people who will propose detailed market analyses, quote Gartner reports, and produce endless slides, presentations, and alignment meetings. At the end of this process, the consultants in suits will blame any failure on the client who did not want to follow their advice all the way through.
In my vision of a skunkworks project, aligning the interests of the client company and the skunks’ team is necessary. The only way to do this is to imagine the latter waiving any markup on the costs incurred in developing the project and agreeing to tie its profit exclusively to the initiative’s success.
In this way, the company would continue to bear the cost of developing the project but would pay the skunks’ team a price similar to the salaries of its most qualified employees.
The skunks’ team would profit from the project only if the product was successful: by getting a percentage of sales for a certain period or acquiring a share of the spinoff for free.
Do you have ideas on how to align the interests of the two parties better? Leave a comment.
Thanks for taking the time to read this episode of my newsletter: I hope I’ve been helpful. If you think my sketchbook might interest someone else, I’d be glad if you shared it on social media and forwarded it to your friends and colleagues.
See you in a couple of weeks 😊
Nicola
This concept is very well codified at Amazon, but we call them "Two Pizza teams". Very interesting topic indeed!
https://polarsquad.com/blog/missapplied-and-misunderstood-2pt-rule
Very valuable Inspiration. Thanks a lot