Exscientia’s $525m Funding Signals Bullish Outlook for AI in Drug Development and Clinical Trials

Publication Date: 17/06/2021

Written by

Imogen Fitt

17th June 2021 – Written by Imogen Fitt – With over 250 unique vendors operating in the market to date and new funding announcements seemingly published every week, investors are clearly confident in the future of AI technology for drug development and clinical trials. This was exemplified with Exscientia’s massive Series D funding round in later April 2021.

Potential investors may be wondering how they too can carve themselves out a slice of this rapidly growing pie. However, most-vendors offerings have yet to be fully proven in the real-world, investment in the sector is a high-stakes game. So how do you pick a winner in this complex sector?

In this article, pre-empting our own updated VC analysis which will be released next month, we examine the top four questions VC investment firms should be asking themselves before digging deep into their pockets.

1. What Stage to invest?

Stepping in later in the game can be costly for investors, as rounds are frequently becoming larger and larger. This is exhibited well in the drug development company Exscientia, whose $26M Series B round in January 2019 is now dwarfed by their Series D investment announced in April 2021.

As stakeholders begin crowding the newly emerging field, many vendors such as the newly emerged Phoremost, are citing oversubscribed rounds. As this happens deal size is noticeably increasing, with the well-known vendor Atomwise announcing an oversubscribed series B investment round in August 2020 worth a shocking $123M.

Entering at the seed stage of funding may seem too risky, considering most companies will take approximately 3-4 years to develop their technology before entering the market, this approach significantly reduces the capital required upfront.

2. What Type of Vendor?

Some sectors of the market are also less “risky” than others. The AI in Drug Development and Clinical Trials Ecosystem is composed of a menagerie of vendors each with a unique USP. These target every phase of the drug development timeline, and typically augment, reinvent or complete upheave current best practice.

At Signify Research, we have characterised these vendors into three types,

Information Engines & Disease Modelling Vendors: which inform target identification and drug development before a candidate is selected.
Drug Design and Optimisation Vendors: who aid in candidate identification and selection.
Preclinical & Clinical Trials Vendors: who speed up and augment the clinical trials process.

These different types of vendors will start to mature at different points. This all comes down to ROI and proof of concept. The simpler it is to quantify and prove an approach works, the quicker the technology is likely to be accepted and adopted, thus generating more revenue quickly.

For this reason, vendors that offer ancillary products which augment singular processes at specific points in the drug development timeline, such as machine learning in medical imaging for clinical trial image analysis, are likely to begin generating significant sustainable revenues more quickly. These products, which are typically marketed by information engine and clinical trials vendors, are further independent in the process, unlike most drug design & optimisation vendors.

Drug design depends, in crux, on a significant number of successful products making it to market. Only then can partners quantifiably evaluate the impact on their business and prove ROI. Once a candidate has been selected, however, this proof can take a long time to come. Clinical trials alone have been known to take six to seven years on average to complete, and with a 90% failure rate[1] that is nothing to scoff at. With no AI-designed drugs available on the market today, and proof only extending to entrance in clinical trials, the true measure of returns achievable for these vendors may be some time away.

Several pharmaceutical vendors are for this reason pursing in-house development to speed up the proof-of-concept and spreading their bets. Whilst investment for in-house development is costly and labour intensive, vendors engaged in drug design who employ this route are much more involved in the development process, better positioned to champion the integration of other AI products, and thus may see returns much more quickly than their independent counterparts.

And whilst investment is riskier, there is no doubt that these drug design and optimisation vendors possess the largest potential for the best returns.

There are several business models offered by vendors in the market today, but by far the most interesting is a risk-sharing model which sees novel drug design vendors initiate deals with mile-stone related payments and royalties upon achievement. Royalties, which offer a percentage of revenue tied to product related sales, are by far the most attractive revenue generating prospects, potentially reaching billions over their lifetimes. With the average patent license lasting on average over ten years (depending on legislation) before expiring, there is potential for vendors to be able to capitalise on a successful deal for some time.

Compared to SaaS and CapEx vendors, who operate on capped revenue models which complete once service is completed, drug design and optimisation vendors start to look increasingly attractive.

3. Where are the experts looking? (The Final Clue)

Finally, the most overlooked clue investors should ask themselves before committing is: who else is eyeing up the company?

Typical VC investment firms are not the only companies shelling out dollar to help bring AI start-ups’ plans into reality. Pharmaceutical giants themselves have begun investing in a multitude of different companies in the market. These giants often possess dedicated venture arms equipped with specialist knowledge of the industry, who are perhaps the best positioned to understand the underlying potential of each company.

Investment from these types of companies is perhaps the ultimate seal of approval, also posing the potential for acquisition, and an easy exit strategy, down the line. Roche’s acquisition of Flatiron for $1.9bn in 2018 is one such example.

Making the Commitment

Despite all these challenges and risks for individual companies, the AI in Drug Development and Clinicals Market without a doubt will remain a lucrative opportunity for investors for some time. In our latest market report, we have forecast the market to reach 3.024bn by 2024, growing at an average CAGR of 34.3% from 2019-2024.

Every segment in this market is set to experience significant growth over the next few years (see Figure 2.1) the sheer number of vendors entering the market is expected to result in a significant number of M&As, bankruptcies, and zombie companies. Whilst not everyone will come out a winner in this rat race, the size of the rewards on offer could well offset the risk of early market entry.

For more information on Signify Research’s market report “AI in Drug Development & Clinical Trials – World – 2020“, or related research, please reach out to the author Imogen.Fitt@signifyresearch.net.