Timing is everything
What does it mean to get it right in deep tech?
Saturday Night Live is back again!
One of the reasons SNL is interesting to me is because of how the show synthesizes standup (in the form of the host’s monologue) and sketches (in most of the rest of the show). Both types of comedy rely upon timing.
Timing is different, but no less critical, in entrepreneurship. Startups, are early, on time, or late not with respect to appointments, but rather with respect to trends.
Founders are often strongly encouraged to think about “Why now?” and include their perspectives in their pitches.
I can say with confidence that most investors I know consider this question, the startup’s answer, and their own view on the startup’s answer as critical data points when they make decisions about whether or not to move forward with a venture.
Early, on time, and late are how they bucket those answers — and those buckets materially affect a fund’s interest in investing.
What’s “on time” for a startup?
One answer is that it’s driven by this technology adoption curve, and where on it the VC fund likes to position its initial investments.
I’m not going to try to express a general rule for where VCs prefer to invest. What I will say is that different VCs definitely have different perspectives on the amount of adoption risk they prefer to take. I’ve seen it vary pretty significantly based on stage, sector, and also geography.
There’s another approach to this, which is more directly focused on the economic potential. I’ve also heard of investors just straight-up working the math on “Will this startup hit my MOIC [Multiple On Invested Capital] target by the time I have to look for an exit?”
VCs may look at both the adoption lifecycle or MOIC math using data, and how startups perform on key metrics is often an indicator as to how interested a venture investor will be in them. Some metrics I’ve seen early-stage founders mention include:
Monthly Active Users
Monthly Recurring Revenue (MRR) or Annually Recurring Revenue (ARR)
Churn
And there’s plenty of others.
Of course, all these things are only useful in evaluating products that have launched.
At Dorm Room Fund, we’re excited to back early-stage student founders before or after launch!
What’s early for some funds will certainly be on time for other funds, and late for yet another group of funds.
Timing in deep tech
These approaches above don’t work so well for deep tech, because early-stage deep tech startups are almost always pre-launch.
The higher switching costs in hardware aren’t just for customers — they also impact the design cycle. This means the OODA Loop for new and improved versions of prototypes or production hardware is necessarily longer.
My view in deep tech is that while the state of startup’s technology (as early, on time, or late) is usually be knowable, the state of the startup’s market often isn’t, because it’s waiting for a key technical innovation. As a result, the answer to this market timing question only really becomes apparent as a later-stage startup, or maybe even the growth equity stage.
The way I see it, “on time” in deep tech means that the product specifications meet buyers’ requirements at the same time as the buyers’ market size expands rapidly.
What this looks like in practice is that the growth of the startup is constrained by their ability to produce the product, rather than the number of potential buyers, or some buyers’ willingness to pay. These startups that are on time cannot build and ship their product quickly enough to service all their customers.
Outstanding success in this is not something that’s shown by anything internal to the firm. The clearest indicator to me that a deep tech startup is outstandingly on time is when their direct competition also has a production capacity problem.
Good examples of this include:
Non-optical earth observation technologies, like Synthetic Aperture Radar, today
Small-lift launch vehicles, in the late 2010s/early 2020s (this is no longer the case since SpaceX has started flying regular Transporter rideshare missions for small satellites)
A solid counterexample would be early electric vehicles. They were so high-end that their ability to enter the market was constrained by retail prices more than production. Now that we’re seeing more auto manufacturers offer budget-friendly EVs, the dynamics are changing — but I wouldn’t describe this category as on time deep tech at this point.
Looking to the future, I think some sectors where deep tech startups have the potential to really get the timing right include:
Robotaxis
Humanoid robots
Quantum computers and their components