During the 2010’s, mobile Internet was experiencing its golden years.
The US started with 20% smartphone penetration and ended with over 70% and 250 million users. Uber roared. Airbnb IPO’ed at a $100+ billion valuation.
But for an investor, this is yesterday’s glory and that means it’s also someone else’s money. So what’s next?
Deep Tech Awakening
Between 2015 and 2018, global Deep Tech investment increased by 22% CAGR. China, the most aggressive of the cohort, saw its Deep Tech investment grow by 80% YoY during the same period.
By 2019, Deep Tech made up more than 20% of global unicorns.
In the same year, Beyond Meat (BYND) put plant-based meat—and the broader Deep Tech category—on the map. Less than three months after its IPO, BYND’s stock soared over 820%, putting its market capitalization at a whopping $13.85 billion. Since then, a barrage of companies have gone on a spree of acquisitions and IPOs.
By 2020, EV SPAC formed an aggregate of over $30 billion in market capitalization, and last month, UiPath—a AI-based robotic process automation company—IPOed at a $29 billion valuation, sealing away any doubt in the future of Deep Tech.
The trend towards Deep Tech is not imminent; it began long ago. What we are seeing today is still only the tip of the iceberg.
But what prompted the shift? Why is it happening now?
In short, several large scale problems have reached a critical point that businesses and societies can no longer ignore.
And these problems can only be solved by Deep Technologies.
Why Deep Tech Now?
Let’s substitute the term Deep Tech with diabetic cure.
Ask yourself, why do we need diabetic cure now?
The answer feels obvious; because 34.2 million – or 10.5% – of the US populations has diabetes, and that number is pre-pandemic.
While mobile internet companies have sunk tens of billions of marketing dollars into creating problems and convincing consumers that life is unbearably difficult when we have to walk two blocks to grab takeout, Deep Tech companies have risen to prominence because consumers and corporations are feeling the increasing pain of real world problems that humans can no longer ignore.
Those problems exist and are felt today, and not in hypothetical situations that may or may never happen:
- The world needs to produce 65% more animal-based protein and double crops by 2050 to feed the planet. There is currently no sustainable solution to meet this demand.
- Aging workers are leaving industries, creating labor shortages in the most labor-intensive and repetitive jobs, as well as opportunities for AI and advanced robotics applications in industrial automation.
- Banks are closing down branches and letting go of security guards while arming themselves with cyber troops as billions of new devices (and security threats) have moved online.
- US Healthcare cost as a percentage of GDP rose from 5% to 18% over the last 60 years, prompting more demands for diagnostics and early detection technologies to curb the long-term socio-economic impact of chronic diseases.
- Problems—not capital—are propelling Deep Tech companies to the front page of the WSJ.
Fortunately for investors, these problems are only going to grow in size and scope. They’ll continue to impact everyone whether they are iOS or Android users.
The software paradigm failed at deal picking
There is little doubt these companies will generate venture-backable returns. The problem is they look extremely different from the mirage of software companies of the recent decades.
In other words, the old playbook no longer works. We need a new way of looking at deals.
The people aspects do not change much; founders will always be founders. We need them to be resilient, relentless, and resourceful just the same, but deep tech investment means we also need them to think about how to build their product and company in an entirely different way.
Product iteration requires letting go of lean and agile startups
Simply put, it is impossible to be lean and agile when each product iteration takes 3-6 months.
Now, I will say that I have a lot of respect for the lean startup and agile development methodology, and to a certain extent the principle still applies.
The methodology aims to shorten product development cycles and rapidly discover if a proposed business model is viable by creating business-hypothesis-driven experimentation, iterate product releases, and validate learnings through customer interactions.
It does not work well when a robotics company might take 3-6 months to build the first viable prototype for field testing.
The napkin plan—which advises drawing your solution on a napkin and talking to customers for insights—will not excite the customer whose only question is the efficacy of the product. It will certainly never land you a meeting with pharmaceutical executives.
Deep Tech founders need to walk into the arena with 10x the confidence that there is a market waiting on the other side, just as we all know there are 30 million people waiting for a diabetic cure. And that is just in the US.
Lock in over traction
Because Deep Tech companies take longer to launch their products, investors are unlikely to see metrics inline with software companies.
VC typically asks for a million dollar annual revenue run rate (ARR) when a company is raising Series A. Some AI companies may be able to achieve this metric by the time they are raising Series A. Others will not, such as battery companies who may not be able to achieve it even after they IPO’ed.
So while there is a large spectrum of Deep Tech companies, another metric that is equally important and in many cases even more important than ARR is the strength of lock-ins.
Battery companies go through years of validation and integration with automotive OEM and Tier 1 suppliers because new vehicle launches are 5-7 years product development cycle. But once they are “in,” it is almost impossible to get them out.
Robotics companies likely sit in the middle of the spectrum. Dishcraft— which collects and delivers “clean plates” to corporate and hospitality industries while automating cleaning using dishwashing robots—chose a tech-enabled service model which will see fast revenue growth because it bypasses the pilot and integration barriers.
On a contract, Picnic—a pizza automation robot—chooses to sell to pizza chains and will have a far stronger lock-in once it is able to land and expand, but its revenue ramp will be slower.
In other words, traction by itself is an absolutely useless metric to Deep Tech companies. Investors must account for the shade of grey and the trade off between revenue ramp and lock-ins.
10x better won’t necessarily convince the customer
VC often praises founders for being passionate about their products and how obsessed their customers are. They are able to create a 10x better “something” that makes their customer’s lives better.
Unfortunately, even the most game changing deep tech products may not necessarily excite the customers.
Deep Tech products often do not touch a single stakeholder. Founders need to understand this, and think through how their products interact with the entire ecosystem, not just the user of the product.
As outlined in my Techcrunch article on the construction industry, general contractors that want to deploy site management tracking tools to improve collaboration and productivity often incur resistance from the many subcontractors engaged for the job.
Once completed, the new project might use an entirely different set of subcontractors. So unless the startup is able to address the dynamics of both the general contractor and subcontractors, the technology is unlikely to see the light of day, regardless of whether it’s a 10x better experience.
Now that we’ve examined a couple of the challenges with applying the software paradigm to assessing Deep Tech companies, it’s time to go deeper.
How would one actually go about picking Deep Tech investments? How do you account for the technology risk? What about the extended sales cycle and the complex market structure Deep Tech companies need to navigate?
Keep an eye out for Part II of the series on how to pick the successes of Deep Tech companies.