If you’ve been reading our blog recently, you may have noticed that we updated our investment thesis and shared that deep tech is on our radar. We followed that post up with what deep tech means to us but since it is a broad category, we’ll be writing more on specific areas within deep tech over time.
With that said, today we’re sharing some considerations we take into account when evaluating a robotics startup. Keep in mind that these are our preferences based on lessons we’ve learned, but that our sample set is small and we’re far from domain experts. In other words, this post is in no means intended as a statement on the right or wrong way to build a robotics company.
1) Vertical integration
We used to say that we’d invest in companies that touch hardware as long as the main value driver was in the software or the data collected. However, this is no longer the case. We’re now comfortable with the idea that hardware doesn’t get commoditized over time. Instead, we think about where value is being created, captured, and accrued… and sometimes, this means that vertical integration is the way to go.
Unsurprisingly, as investors, we always think about defensibility. As we previously wrote, the path to defensibility differs for software and deep tech companies. Hardware is hard to start up and scale; but once you scale, you have the opportunity to build a highly defensible company. While for software companies, starting up is easy but defensibility is hard (unless there are true network effects).
Of course, scaling hardware and vertical integration (when the focus is not only software) can be quite capital intensive given all the R&D, manufacturing, supply chain logistics, implementation, and professional services involved. Companies need to ask what hardware to build and what to buy off the shelf.
For example, Pickle Robot (a V1 portfolio company) uses the Kuka robot arm and built specific hardware / IP around gripping and the chassis that moves the robot in, out of and within the truck to unload packages. A yet-to-be-announced portfolio company, who traverses extreme terrains for data collection, built the chassis. Both companies develop the AI/autonomy in house.
TL; DR. Deciding whether your start-up should be vertically integrated isn’t straightforward but should be guided by maximum value capture to get to defensibility and scale over time.
2) Vertical vs Horizontal
We skew towards a killer use case first with the opportunity to be a platform versus starting off as a platform with the potential of endless possibilities and functions. That means we like seeing demand for a product that solves a specific problem instead of building a tech solution in search of a pain point. While we appreciate R&D and academic research, our primary focus is the market need.
From what we’ve seen to date, horizontal plays (e.g. generalized robotics/physical AI) may require more co-development or a tight design partnership from customers, and likely more customization and professional services, especially in the early days of looking for product market fit when solutions seem more bespoke. Additionally, the go-to-market plan for a horizontal company can be tricky, since the options are so broad. And companies need to decide if they’ll go to market with system integrators or go it alone.
TL; DR. Solving a specific problem for a specific customer first may be the best way to start up. It’s important to be good at one thing to gain scale and grow quickly. The caveat, of course, is that there is a risk that a customer’s problem is too specific or narrow which might limit potential upside… but hopefully, what looks niche at first can translate into a big opportunity down the road when you can expand into adjacent markets, etc.
3) Understanding the purpose of automation
What’s really at stake when it comes to automation? Are robots going to be cheaper than labor? What happens if the labor involved is genuinely cheap?
A robotics company needs to understand the margins of the target industry/use case to know if there’s a strong enough value proposition for automation to address workforce shortages and/or churn. Ideally, the price point for a robot is a no brainer, so there’s no comparison to labor. If the price of automation and labor are comparable, customers might be more reluctant to invest in the “new thing.”
For example, let’s say you’re building a sorting machine and it’s priced at the same or slightly less than a human sorter’s salary. It wouldn’t be surprising for customers to evaluate your technology on “sorts per minute or hour.” But, if your sorting machine is priced much cheaper, customers will be more forgiving if the machine is a little slower at the beginning. It would be worth it to try and see if the machine gets faster over time.
Another important factor is what’s at stake for the customer? What’s the cost of a mistake by a robot? Is the risk of getting it wrong higher than not? This is similar to what we wrote about back in 2016 with trust and AI. Robots get smarter when mistakes can be made, so are you able to factor in the potential for mistakes when building it? There’s typically more room for error in consumer solutions versus the enterprise.
TL; DR. It’s trendy to “automate all things,” but is it worth it?
4) What’s the business model?
A while back, Boris talked about how AI can spark both technological and business model innovations. More specifically, transformational change occurs when there’s a disruption in both technology and business models.
We constantly hear about how the future of AI will be about “selling work and not software.” We see the same trend in robotics…e.g. selling “work” and Robot-as-a-Service (RaaS). Everyday, RaaS is becoming a more accepted business model in the enterprise, but there are still many industries that push back on OpEx vs. CapEx, with a preference toward purchasing the robots/hardware outright to gain tax/depreciation benefits.
TL;DR. Business models for robotics companies will vary, not just from industry to industry, but also across companies within the same industry depending on their priorities.
In sum, it’s becoming increasingly cheaper to build robots thanks to AI, 3D printing, and more flexible manufacturing and supply chains. We are excited to meet entrepreneurs who are automating enterprise and consumer pain points. Please reach out!