AI startups are taking the industry by storm – there is no dearth of early stage, vertical AI startups that have put artificial intelligence and deep learning as the core of their value proposition. But how can venture capitalists and tech savvy investors who may have only a baseline understanding of this emerging technology make good business decisions?
By and large, there are six core competencies that investors look out for, UK -based MMC Ventures notes — a) strategy, b) technology, c) data, d) people, e) execution and f) capital. However, in cases related to AI/DL startups, it goes beyond Proof-of-Concepts/narrow AI prototypes and technical feasibility to assessing whether a startup can grow at scale and how it overcomes the persistent black box problem of AI.
Over 100,000 people subscribe to our newsletter.
See stories of Analytics and AI in your inbox.
New York based CB Insights that covers the startup ecosystem has assiduously captured the rush to gobble up vertical AI startups that started in 2010 with Apple taking over Siri. Post 2010, 2015 and 2017 saw a feverish activity in the M&A space – the race to acquire AI startups is dominated by Google, Apple & Facebook. It also regularly chronicles the rise of enterprise AI startups ripe for investment, comes up with well-funded categories which serves as a primer for VCs, VC Funding trends and even publishes VC rankings.
AI Investment Landscape
According to McKinsey’s research, tech giants spent $20 billion to $30 billion on artificial intelligence in 2016, and 90 percent of this spent was on R&D while 10 percent was on AI acquisitions. VC and PE financing, grants, and seed investments also grew rapidly, from a small base, to a combined total of $6 billion to $9 billion, cited the McKinsey research.
Machine learning seen as an enabling technology, has attracted 60 per cent of investment – since it is seen as an enabler of other technologies and applications such as robotics, speech recognition. Besides machine learning – other areas that have drawn the most investment are Computer Vision, NLP, Autonomous Vehicles, smart robotics and virtual agents. Even though the buzz around AI has grown louder, investment in AI is still in the early stages and makes up a small fraction – 1 to 3 percent of all investment by PE firms, outlines the research.
However, there are few things which are certain about the AI landscape:
- Over the last two years, investors have grown more confident about AI & its potential
- Which means the deals are only getting bigger in size (AI investments are usually deemed to be cash-intensive)
- The areas attracting most investment after machine learning are Computer Vision, NLP, driverless tech & virtual agents
- A statement that often comes up is that investment in AI is akin to a marathon and takes a long time to pay off.
- A key hurdle is that even though the market for AI adopters is growing, AI is not yet integrated in business’s value chains at scale and a lot of enterprises are still in pilot phases. This could lead to a slowdown in the uptake of AI at scale.
So, how can VCs look for winning technologies? We list down a few metrics:
A visionary team: Successful AI-driven startups will be led by a strong CTO and an even stronger, visionary CEO, and it is here that a strong technical expertise will come into sharp focus, noted Bonnie Foley-Wong, CEO of Pique Ventures and author of Integrated Investing. CEO and CTOs are the public face and it’s their product vision that guides the startup right from an stage to development phase and the time of going public.
Data is the new war chest: According to Yuval Ariav, data science entrepreneur and an active investor, given how every startup is being labeled as an AI startup, VCs and tech savvy investors have to work even harder to separate the winners from the losers, or in Ariav’s words “the signal from the noise”. This means that data forms a crucial investment and AI-based startups should demonstrate a way to generate a data cycle, or much better have access to proprietary data – only then it would be “investable”.
IP increases the likelihood for an investment: According to AI advisor Francesco Corea, what makes a startup a viable investment option – is if they have already filed patent. Most AI startups say their technology is patent pending which puts them in an advantageous position and gives an edge over the competition – a good patent portfolio can make the startup more valuable to potential investors and also makes investors more confident about the technology.
However, the fact is that anybody can file a patent application, and a patent filing won’t make the startup’s technology any more valuable, but startup founders believe it increases the startup’s valuation. So, what’s in it for a VC – in case of exits, patents are sold off to get the return on investment and in case of bigger exits, a patent can increase the startup’s valuation.
Teams with technical advisors are well-positioned for attracting further investment: This is a common phenomenon – most noteworthy AI startups are backed by veteran technical advisors or industry experts who play a huge role in steering the startup on the business side effectively, act as mentors and also help raise external capital. A technical advisor who sits on the board raises the profile of the startup, helps steer the company’s vision and also makes it a viable option for investment. Case in point – Bangalore based DataVal Analytics is helmed by technologist and telecom guru Dr Sam Pitroda who is onboard as a Chairman & Co-founder.
Over the years, the AI Venture landscape has become extremely competitive space with big tech giants and VCs adding an AI component to their investment portfolio. Leading companies like Google, Apple and Facebook acquire startups to shore up their talent base instead of building a team from scratch.
Some of the caution areas for investors are:
- When it comes to AI startups, it is hard to come up with a PoC, and a narrow AI prototype won’t be enough to get the market validation. In this case, VCs usually go for startups that have an MVP backed by the team’s technical expertise and a virtuous data cycle.
- Black box problem of AI is what holds back most investors from parking their money in potential startups.
- Besides technical expertise, startups should have a business case since it gives a proven solution and adds more weightage rather than just banking on tech
- Another key hurdle for investors would be identifying the growing categories in AI space, which are ripe for investment. For example, CB Insights identified enterprise AI startups emerging as the most well-funded categories and startups are offering services ranging from automated data collection from unstructured sources to even real-time data analysis.
- There was time when chatbot market was booming but with a proliferation of startups, investors are steering clear of bot startups.