What is Deep Tech?
At its core, Deep Tech is the business of turning frontier science into products, the commercialisation of laboratory breakthroughs through structured technology-transfer processes. Whether the advance comes from a new material, a novel bioprocess, or a cutting-edge AI algorithm, the venture’s value proposition hinges on owning (or licensing) that underlying intellectual property and engineering it into a scalable solution that can redefine an industry.
# Five hallmarks
These technologies share five hallmarks, as noted in previous literature:
- Radical novelty, they introduce genuinely new technical capabilities.
- Fast growth, research, patents and investment expand rapidly.
- Coherence, a recognizable knowledge base and community forms around them.
- High-impact potential, often with long development timelines, their main socioeconomic effects are expected, not fully realised.
- Uncertainty & ambiguity, outcomes, markets and standards are still fluid.
By contrast, sectors such as fintech or general consumer software build mainly on technologies that already exist. They remix mature stacks of code and cloud infrastructure to create new user experiences or business models, but do not introduce a fundamentally new technology layer. In that sense they are sometimes dubbed "shallow tech" or "regular tech", innovations driven more by application design than by scientific discovery.
In Latin America, Deep Tech ventures are commonly labelled Emprendimientos de Base Científico-Tecnológica (EBCT) amongst policy papers, innovation-law frameworks and regional multilaterals like ECLAC. In China, we have found that the homologous term is "New Infrastructure," embedded in State Council guidelines and provincial five-year plans as the umbrella for AI, advanced manufacturing, and other Deep Tech priorities.
| Category | Definition | Example use case | Value of the use case |
|---|---|---|---|
| Artificial Intelligence | The ability of a machine or computer system to perform cognitive functions usually associated with humans, such as learning, problem-solving, and decision-making. | In healthcare, improving diagnosis accuracy with medical-imaging devices (e.g. Samsung’s ultrasound-based tool, which raised breast-lesion diagnosis accuracy by ~5%). | Enhances the customer relationship, better deals, virtual product trials, subscription management, payment verification, done in a privacy-friendly way. |
| Advanced Materials | Technology that modifies or engineers materials at a molecular level to create superior properties: enhanced durability, conductivity, or sustainability. | Environmentally friendly batteries for electric vehicles (e.g. Tesla, General Motors), storing and releasing energy more efficiently. | Supports climate and eco-friendly transformations; foundational to renewable energy systems and lightweight structures. |
| Biotechnology | The use of living organisms or biological systems to create products, technologies, or processes that benefit humanity in health, agriculture, and energy. | Liquid biofuels and methane from organic waste (e.g. ENI, ExxonMobil), using microorganisms to transform waste into energy. | Improves sustainability; addresses global health through new medicines and food scarcity through enhanced agricultural products. |
| Blockchain | A decentralized, immutable ledger technology that stores transactions and tracks assets, enabling secure, transparent records without a central authority. | Tracking loan conditions between borrowers and lenders end-to-end (e.g. Indra borrowing €75 million from BBVA). | Enhances tracking, traceability and safety in distribution; greater transparency in negotiations and supply chains. |
| Robotics and Drones | The use of machines to perform automated tasks, often in complex or hazardous environments; includes flying craft and remote-controlled systems. | Automated store systems retrieving products from shelves (e.g. Amazon Robotics); drones for item pre-sequencing or last-mile delivery. | Speeds up traditional processes such as pallet movement and barcode scans; improves order accuracy and shortens wait times. |
| Photonics and Electronics | Technology that enhances the properties of photons (light) to transmit information or actuate; a similar process with electrons in electronic devices. | Advanced cameras, thermal or multi-spectral sensors for food safety (e.g. PepsiCo’s vision-inspection systems detecting potato defects). | Increases transparency and customer trust; critical for telecommunications, computing, and medical diagnostics. |
| Quantum Computing | Another way of processing information, leveraging the properties of matter at a nanoscale to perform calculations impossible for classical computers. | Optimizing radio cells in telecommunications (e.g. operator TIM, using quantum principles with hardware maker D-Wave). | Reliable, high-performance mobile services; potential to solve complex problems in medicine, finance, and materials science. |
Source: IESE, How Corporate Giants Can Better Collaborate with Deep-Tech Start-ups
# Deep Tech investment: demystifying the risk
Beyond defining Deep Tech itself, the primary hesitation surrounding investment often emerges from evaluating critical metrics, particularly funding cycles and return on investment, in comparison to traditional technology sectors such as fintech, where conventional tech typically demonstrates immediate and steady growth upon receiving capital. This concern can be especially pronounced in LATAM, where investors are familiar with shorter business cycles and may be reluctant to wait several years for tangible returns.
Fortunately, various studies have started to quantify these concerns. A McKinsey report analyzing European and U.S. Deep Tech funds reveals an average net internal rate of return (IRR) of 17%, outperforming traditional tech funds, which yield an average of 10%. However, the absence of directly comparable LATAM data complicates contextualizing these findings, underscoring the need for localized research. The 2025 European Deep Tech Report identified and addressed eight prevalent misconceptions about Deep Tech investments.
| Question | Verdict | Detail |
|---|---|---|
| Is Deep Tech an undefined term? | No | Deep Tech is novel science being shipped in a first-of-a-kind product. |
| Is Deep Tech a new phenomenon in venture capital? | No | Deep Tech venture financing has always existed and shaped the sovereignty of nations. |
| Do Deep Tech companies need more capital? | Yes | But often money is spent on building moats. |
| Do Deep Tech companies take longer to achieve revenue? | Yes & no | True in the early years, false at later stages. |
| Do Deep Tech companies fail more often? | No | Similar failure rates compared to regular tech. |
| Do Deep Tech companies need more time to exit? | No | Similar exit timelines compared to regular tech. |
| Do Deep Tech companies have larger exits? | Inconclusive | Some large outcomes exist; Europe needs more big exits. |
| Does Deep Tech investing deliver top returns? | Yes | Some data shows higher IRRs than regular tech. |
Source: Hello Tomorrow, The 2025 European Deep Tech Report
While precise metrics for Latin America’s Deep Tech ecosystem, and granular KPIs for each vertical, are still under development, a high-level risk-profile comparison helps illustrate the fundamental differences. Deep Tech startups carry substantially higher technology risk, require large upfront capital, and face protracted development cycles measured in years rather than months. Yet, once past those hurdles, they enjoy a powerful competitive moat, anchored in proprietary scientific breakthroughs, extensive IP portfolios, and teams of niche technical experts, that most "regular" tech firms simply cannot replicate.