When, why, and how to invest in research within your startup


Wondering how to build research into your startup? Find out when to invest in research, how to motivate researchers and the best ways to encourage collaboration so you can drive long-term innovation.
Research can be a powerful catalyst for innovation and growth within a startup, but building an effective research function can be challenging. It requires navigating competing priorities, departmental differences and implementation problems — all while trying to stay ahead of the curve in your market.

I’m a scientist and the founder of Pinecone, and I've been through the ups and downs of building research teams and integrating them into broader organizations. In this post, I'll share some of the key insights I've gained along the way, from knowing when to invest in research to keeping the peace between scientists and engineering teams. 

First, though, let’s talk about a model that I believe works really well.

Success secret from a scientist CEO: Build an immersion model

In my experience as a scientist and CEO, I've found that one of the most effective ways to ensure the success of a research team is to implement what I call an "immersion model." We used this model at NPS, and I found it worked very well. We use it at Pinecone, too.

The key to the immersion model is dedicating a small team of one to three scientists (along with a product manager and an engineer) to a single, critical project within the company. Rather than having researchers float between different teams or work on multiple projects simultaneously, they are fully embedded within one project and accountable for its overall success.

This means that the scientists are not just responsible for their individual research tasks, but for the project as a whole. They work closely with the other team members to ensure that the research is aligned with the project's goals and that the insights generated are effectively integrated into the product.

For the immersion model to work, scientists must be truly embedded within the project full time. They should be deeply familiar with the project’s engineering challenges and product requirements.

Typically, a researcher will remain embedded within a project until it reaches a certain level of maturity or their specific research objectives have been achieved. At that point, they can transition to a new project where their skills and expertise are needed.

When to build your research function

The timing of when to establish a research function within a startup can vary depending on your company's focus and goals. 

For some companies — particularly if what you do is solve problems or push boundaries in areas where you’re not sure if progress is possible — building a research team should be an immediate priority. This was the case for my company, Pinecone, where a core part of our mission is to tackle challenges that may not have clear solutions.

However, for the majority of startups, the need for a dedicated research function may not be immediate.

That said, both under-investing and over-investing in research can pose risks to a startup. For companies operating in deep tech, AI, data and infrastructure spaces, failing to innovate can be an existential threat. In these fields, there's always the danger of being disrupted by a competitor who develops something exponentially better, faster or easier to use. If that’s the case for your startup, having a research function in place relatively quickly is important to ensure your company isn't left behind.

On the flip side, pouring too many resources into research too soon can also be detrimental, especially if your startup is still in the early stages. It's crucial to strike a balance and allocate your research budget wisely based on your company's specific needs and risk profile.

Keep in mind that even if you establish a research team early on, their work likely won't significantly impact your first product. Research is typically focused on longer-term innovations that may not have a big impact until one to three years down the line. Make sure you set up your research function with realistic expectations about timelines.

Motivating your research team

Once you've assembled the right group of scientists for your research team, you will probably discover that traditional forms of motivation aren't necessarily required. Good scientists are often internally driven — fueled by a deep passion for their work and a desire to push the boundaries of knowledge in their field.

In many ways, scientists in a startup are similar to tenured professors in academia. Many professors work long hours well into their later years, even after they achieve tenure, simply because they are motivated by their research. The same holds true for scientists in the business world — they are driven by a love for what they do and a desire to make an impact.

That said, there are certainly ways to inadvertently demotivate your research team. One major demotivator is making scientists feel irrelevant or disconnected from the company's goals and mission. If researchers feel their work isn't valued or isn't contributing to the bigger picture, they can quickly lose motivation.

Another potential demotivator is bogging scientists down with excessive execution tasks. While some level of process is necessary for any team, over-burdening researchers with administrative work or tying them up in red tape can be incredibly frustrating and inefficient. Scientists need the freedom and flexibility to practice their craft and pursue the research questions that drive them.

Avoiding cultural gaps between research and engineering groups

When you have both research and engineering teams within your organization, there's always a risk of cultural gaps forming between the two groups. These gaps can lead to tension, resentment and a lack of collaboration, which can hold back your progress. As a leader, it's important to proactively address this issue and intentionally create a culture of mutual understanding and support.

One key way to head off cultural problems is to clearly communicate your expectations for how the research and engineering teams should interact and support each other. Make it clear that both groups are working towards the same overall goals, even if their day-to-day work and priorities differ.

Another important step is to help each team understand and empathize with the other's perspective. It's easy for engineers to look at researchers and think they have it easy — running experiments and writing papers while the engineering team is working long hours to meet product deadlines. Researchers work long hours, too, but they are focused on different things than the engineering team.

To bridge the gap, highlight each role's unique challenges and rewards. Help engineers see that researchers are also working hard, even if their work is not immediately visible. Encourage researchers to appreciate the critical role that engineers play in bringing their ideas to life and ensuring the product's stability and performance.

Emphasize that each team has chosen their particular role because it aligns with their skills, interests and passions. An engineer who loves building and operating complex systems wouldn’t be happy in a research role, and vice versa. Point this out to your teams to help prevent resentment and "grass is greener" thinking.

The balancing act of startup research

Integrating research into a startup can be a powerful way to drive innovation, but it requires navigating a complex set of challenges and trade-offs. By being strategic about when to invest in research and promoting collaboration between teams, you can set your research function up for success.

If you’re building a research team at your startup, here are the top things to remember:
  • Implement an immersion model to integrate researchers into product teams and ensure alignment with company objectives.
  • Carefully consider when to invest in building a research function based on your company's specific needs and goals.
  • Recognize that research may not have an immediate impact on your first product, but can drive innovation in the long term.
  • Motivate research teams by eliminating red tape and providing an environment that supports their passion for pushing boundaries.
  • Actively work to bridge cultural gaps between research and engineering through clear communication and empathy.
Have any ideas or suggestions to improve this article?