What is deep tech? Life after consumer apps

Deep tech is a phrase describing organizations (often startups) that depart from end-user products or services to focus on technologies that require significant technical or scientific advances. The phrase deep tech marks a distinction from the consumer-facing applications that most people associate with the tech industry. So, an “Uber for X” app that lets you buy something through your phone isn’t deep tech, no matter how innovative or profitable it is. A new form of material science that puts faster or cheaper chips into phones would be deep tech, as would be innovations in cellular broadcasting to provide those phones with a faster or more reliable Internet connection.

The phrase was coined in 2014 by Swati Chaturvedi, founder and CEO of the online investment platform Propel(x). In a 2015 LinkedIn post, she lays out her vision for the phrase, explaining that it offered a way to distinguish “startups in the life sciences, energy, clean technology, computer sciences, materials, and chemicals sectors” from the “unicorn” startups that so much venture capital was chasing in the mid-2010s. Propel(x) was founded to draw angel and venture funding to deep tech companies, so it’s still a phrase that’s part of the business side of the industry rather than a technical term in the strictest sense. But as the shine wears off the hot web and mobile startups of the last decade, there is more and more interest in companies looking to make advances in deep tech.

What counts as deep tech?

Propel(x) may have popularized the phrase, but deep tech has now taken on a life of its own in tech and venture-funding circles. As a result, there’s no centralized authority that can declare what counts as deep tech and what doesn’t, and as the phrase carries gravitas and the potential for funding, some take the opportunity to stretch its meaning. Still, there’s a certain set of technologies that consistently makes the cut for being known as deep tech:

Elements and goals of deep tech

What do all these technological areas have in common? There are a few common threads that tie them together:

• Solutions that aim to overcome physical challenges: For consumer-facing apps, much of the innovation goes into providing frictionless connections between customers and already existing businesses and resources, and taking a cut of the resulting efficiencies. Deep tech startups, by contrast, face challenges at the level of physical reality, rather than human institutions and networks. Creating new medicines, or room temperature superconductors, or using quantum physics to create new computing paradigms—these are the realms where deep tech startups work.
• Combining divergent technologies into a larger solutions: Tesla and other electric car manufacturers need to bridge several disciplines when building their vehicles, combining computer science and chemical engineering to produce smaller, cheaper batteries that can still power a car that people want to drive. In another example, you might seek to harness the power of AI to discover new cures for diseases. These are the sorts of challenges that deep tech startups face.
• Solving big picture problems: Hopefully it’s clear by this point that deep tech startups often look to tackle fundamental problems facing the human race. No offense to pizza delivery apps (because who doesn’t love a pizza), but deep tech companies are looking to apply their lessons to issues like disease or climate change, or to promote big leaps forwards in computational power or manufacturing processes.

One thing to note here is that while many of these fields are in what we might consider the “traditional” realm of the tech industry, some of them go beyond it. As Chaturvedi said in an open letter in 2021:

Our worry is that most people still default to thinking of deep technology in the realm of information technology and computer sciences alone. Only in passing do people also mention innovations in life sciences or industrial technologies and others in the same breath. As a result of that narrative, Venture Capital has become solely focused on all things computing (“AI for everything” is the byword these days). But other meaningful deep technologies, that have the potential to change the world, still do not get much financing interest … just a few years back, [Tesla] had real difficulty fund-raising and was on life support with government loans and Elon Musk’s own money. Tesla is an electric vehicle, the cornerstone of which is a battery technology, supported of course by smart battery management systems. But fundamentally, it is chemistry. We need to champion that. We need to champion the disciplines of technology that help us step into the future.

What’s different about deep tech?

In order to achieve the sorts of goals just outlined, deep tech companies have different needs and business processes than customer-facing businesses. That’s why the category was developed within the venture capital community in the first place: because potential investors need to understand that deep tech companies require larger up-front investments and a longer runway until you can expect profitability or an exit event like an IPO or acquisition.

Some practical ways that deep tech companies conduct themselves differently include:

• Moving deliberately and treating things carefully: For much of the early twenty-first century, tech startups followed the Facebook mantra of “move fast and break things”—in other words, make lots of small and iterative changes to your product or platform to introduce new features and push the state of the art, even if that means some things occasionally go awry in the process. (In the software world, this philosophy finds expression in CI/CD and devops.) Along the way, companies may discover that what they thought was their main selling point was really a sideline, and pivot to another service or strategy. This attitude has become so ingrained in tech business culture that it’s easy to forget that it has historically not been the norm. Moreover, when it comes to long-horizon projects undertaken by deep tech companies, it’s not a viable strategy; the innovations these companies pursue can’t be released half baked, because they face higher regulatory hurdles and more stringent safety demands.
• It’s a process, not a product: Software can of course be cooked up by anyone with a computer, and even most physical products can take advantage of existing factories and supply chains once a prototype has been developed. But by their nature, many deep tech products require substantial investment even after the research and development phase is over: they may require specialized factories or whole new supply chains to be assembled before they can be profitably produced as a salable product.
• Plugging into a wider ecosystem: Angel investors or VCs are ultimately interested in profiting from a successful product launch as a result of their deep tech investments. But because of the long timeframes and risks of failure inherent to these sorts of efforts, for-profit entities can’t be the only players here. Many deep tech innovations arise from universities and government-funded labs; others bubble up from “skunkworks” divisions of big engineering firms, where scientists are given a much longer leash to pursue interesting research than is typical at a corporation. This ecosystem is crucial for deep tech success, but it can also complicate the process of monetizing the end results of the research—universities may require ownership of patents, for instance.

Challenges for deep tech

One of the biggest challenges faced by deep tech startups should be clear by the points in the last section. They take a long time to produce profitable products. Indeed, because the innovations they’re pursuing are so far ahead of the current state of the art, they may never produce anything that can be offered for sale, let alone earn a profit.

When deep tech companies do hit the point of releasing a product, they may find that the internal organization that brought them R&D success isn’t suited for this new phase. This is an area where investors with experience working with other companies can really offer expertise in restructuring or bringing in new leadership that can turn an innovative prototype into something that can be mass manufactured and offered to customers.

What deep tech companies absolutely don’t want to do is make promises they can’t keep, or blur the line between the innovation that should be their focus and more prosaic services. Perhaps the ultimate deep tech cautionary example is Theranos, the infamous health tech startup that claimed to be working on rapid blood tests that only required a single drop of blood to work with. Theranos went to market with lab equipment that was built from off-the-shelf parts while its deep tech research foundered, ultimately leading to corporate collapse and jail time for its founders.

Deep tech companies

To give you a sense of what the current deep tech landscape is like, here are some current active startups that have earned the deep tech label from various observers:

• LabGenius: A biopharma company using machine learning to develop protein therapeutics.
• BotsAndUs: Developing autonomous, AI-driven robots that can provide insights into warehouse operations.
• Flexciton: Seeking to streamline the chipmaking process.
• Gourmey: Creating sustainable, lab-cultured meat.
• Deep Vision: Conducting real-time video analytics and natural language processing with a custom-designed chip.
• AgNext: Providing food quality assessment through a mix of AI, ML, IoT devices, and data analytics.

Is your startup destined for this list? Or are you interested in investing the next big deep tech thing? Hopefully, this article has given you a view of the landscape, so you can start your journey.