Deep-tech technologies or Deeptech are described as innovative solutions designed based on distinctive and hard-to-reproduce advancements in Science and Technology. These approaches have strong research support as a baseline and tend to either disrupt Industries or create new markets.
In the late 20th century, the world witnessed an upsurge of ICT with digital platforms and apps doing the rounds. The plethora of these technologies led to a diminishing level of innovation and seized to see any further spike.
Remember Zero to One? So, after the buzz reaching an extent, the prevailing ICT jumped from 0 to 0.1 but not to 1. So what next?
Let us take an example of a ride-sharing app. The app offers a powerful digital platform to find real-time options for ridesharing, carpooling and bike pooling be it as a ride giver or a ride taker. The value offered is apparent but up to a point. Sooner or later there will be competitors strategizing to get the major chunk of the market and the customer would have the choice to opt for the best and most convenient.
How would the app manage to get more users? Or how can one innovate to offer more value?
Now suppose the app makers partners with a product intelligence platform that provides that understand user-intent, generate insights through analytics and deliver personalized experiences. The app leverages the user’s behavioural insights and rolls out a feature that enables the user to pick/offer a ride from anywhere throughout the ride. This would help the app to boost its user base and also feature adoption.
Instances like these are many and indicate a more effective diffusion of technology. In our example, the app is the tech and the insights generated is Deeptech.
The above example is taken from a customer case study of Apxor – our portfolio startup.
Building a Deeptech startup
Over the years, building a tech startup has increasingly become easier and efficient with the internet flooded with different playbooks, guides and methodologies such as Lean Startup which was initially made famous by Eric Ries for software startups is now applicable for almost all startups. Startups today can start, experiment, iterate, learn and scale their solutions much faster at lower costs.
But when it comes to Deeptech startups, founders need a different set of thinking and attitude towards solving the problem. It is because either the problem is too complex and you need to think through or the technology is not ripe enough to implement as a solution.
“Running early stage deep tech start-ups, especially AI, when compared to traditional enterprise software involves significant customer-specific deployment efforts beyond the support/success functions, thus requiring a service+product mindset, which differs from the general expectation that it is a pure product play” – Nayan, Founder Answerwise
The problem identification is rather easy but when developing the solution, it takes months and even years to build the first cut, let alone the MVP. Since most of the problems at hand are of a larger magnitude: think Cancer or depleting natural resources, there are distinctive and ample needs or challenges to build the solutions.
In a survey conducted by Hello Tomorrow and BCG, and answered by more than 400 deep-tech startups, the challenges that respondents identified most frequently included lengthy time-to-market (27%), high capital intensity (25%), technology risk and complexity (17%), and yet-to-be-developed commercial applications (14%). These challenges have been covered in a previous post and can be referred to here
To address the challenges, deep-tech startups need to go beyond funding (which 80% of the startups surveyed ranked among the top three challenges they faced) to such issues as market access (61%), technical expertise (39%), and business expertise (26%).
The needs and Challenges on the way
“True AI and ML capability building takes lot of training data, expert resources and lot of patience. Startups need to keep this in mind before venturing into starting a deeptech company” – Chaitanya, Founder Etta.ai
Research Collaboration
Research lies at heart for Deeptech startups. Arguably, a Startup’s dependence on research for building solutions through deep technologies is around 70-80%. Startups collaborate with Universities and Research Institutes to aid their tech building process through Incubators and programs offering such collaboration.
Challenges – Time-consuming process, needs consistent direction and relevance to the problem, gaps in knowledge transfer and different priorities; although shifting slowly, Commercialization is never a priority for academic research.
High on funding requirements
From experimentation to prototyping (extensive data requirement in case of AI/ML) to testing, especially in case of Healthcare startups -clinical trials and testing on expensive equipment demands a heavy investment even at an early stage.
Challenges – Early bet on technology, trust in Deeptech solutions, higher Time-to-market and often blurred traction for the investment leads to lower willingness in investors to put money.
TRLs – Technology Readiness Levels
Deeptech startups often leverage technologies at a low Technology Readiness Level which is a crucial factor towards the risk and complexity in deploying it as a solution.
Technology Readiness levels indicate the status of the technology from an early stage of evolution when there is just a hypothesis to fully developed product tested in an actual environment.
Challenges – The lower the TRL is, the higher is the risk and more distant it is from being deployed in the market.
Market readiness
Market readiness shows how close the product or technology is to commercial application and how mature its existing customer base is. Take the case of AR/VR in India. The technology adoption has diminished since it was first launched to a mesmerised audience.
Challenges – Lower market readiness indicates a lower customer base and non-existent market (maybe based upon Industry trends and heavy regulations) and complementary infrastructure(Classic example is Apple) which means either it will take more time for market adoption or it doesn’t offer enough value to customers. In both cases, it is a dead-end for startups.
A well-functioning ecosystem
As encountered in most of the above pointers, an underlying need for resources is common to Deeptech startups. Resources for building the product, testing it, taking it to the market and then scaling it to a larger customer base. For an early venture, doing all the activities in-house is difficult and hence Deeptech startups rely largely on stakeholders that help with the Startups’ specific needs at different stages. These stakeholders are people, companies, infrastructure providers, and government-linked through informal and formal networks.
Challenges – There is a dearth of Strategic collaborations helping Deeptech startups in their journey. The ecosystem is evolving but the depth of engagement needs a deeper analysis to identify gaps and opportunities.
Concisely, Deeptech Startup founders need to persevere, be consistent and intuitive to develop solutions and capture value. These solutions take time through research to be built as a robust one, need to have higher TRLs and Market readiness to be deployed as commercial products and seek funding from private and public sectors. To enable all this, it takes an ecosystem from R&D to the first customer and several actors supporting an early-stage Deeptech Startup at different stages.
Insightful? Watch this space for more such takeaways ↑→ https://cie.iiit.ac.in/startup-jukebox/
Until next time!
– Sunita Kumari, Team CIE
