This is a guest post by Richard Armstrong who is an early stage investor in many startups in both SE Asia and globally. Plus Richard has cofounded several companies.
Guest Author: Richard Armstrong
Exits are the name of the game in venture capital, particularly in tech. But for the past couple of years M&A has been very soft.
According to PwC, from 2021-2 the value of tech M&A deals dropped by 40-50% and the number of deals dropped by 30-35%.
And from 2022-3 the value of tech M&A deals dropped by 35-40% and the number of deals dropped by 25- 30%.
This is a very significant drop and it makes venture capital as a whole a less attractive investment vehicle.
Which then begs the question… why is VC money still so hot in this AI field? It seems like tap is still flowing very heavy in AI with all kinds of companies getting funded.
So I was pretty interested when I started to hear about this new way of doing M&A in AI as it makes a ton of sense from a number of perspectives.
What is this ‘new approach’ to M&A in AI?
Well it’s pretty simple. It is a lot like the old acquihire model in that instead of buying the company, you just hire the founders and key employees from the AI startup.
Plus you license their technology.
And the ‘earn out’ terms that you would have typically put into the acquisition terms, you instead put into the licensing terms.
This gives you the best of both worlds:
- You reward & incentivize the founders similar to the way they would have had you acquired them.
- You do this without drawing attention from the antitrust regulators and the transaction costs are often far less.
Let’s run through a few examples so you see how it works.
Example 1: Inflection AI
Microsoft hired most of the key employees of Inflection Ai in early 2024, including the co-founders.
They also licensed their models for $620m, paid out over time of course.
Additionally, as they ended up hiring most of their 70 employees, they paid $30m to waive legal rights related to the mass hiring.
The deal avoided the attention of the regulators and everyone was happy.
Example 2: Adept AI
Adept AI is developing general intelligence systems that can understand and execute complex tasks on a computer.
The company’s primary product is an AI-powered tool that acts like a digital assistant, capable of understanding natural language instructions and performing tasks across various software applications, such as navigating websites, using software tools, and managing workflows.
Amazon hired the key employees of Adept AI in early 2024. In total they hired about 2/3 of the employees and obtained a non-exclusive license to the technology, including the AI models and datasets.
This gave Amazon the technologies they needed while still leaving the company to operate under a new CEO with a much smaller team.
So it is that they’re just trying to bypass regulation?
In a way, yes. And as it turns out US regulators are not completely oblivious to this.
As they are looking into the Microsoft-Infection AI deal to determine if it was structured specifically to avoid regulation.
And similar has happened with UK’s Competition and Markets Authority investigating a deal between Microsoft and French AI startup Mistral. Plus Amazon’s investment in US AI firm, Anthropic, is currently also being looked into.
So i’ll be keeping a watchful eye on how that evolves, but so far I think all these deals should pass through.
Why am I excited about this new form of M&A?
Well as an investor I still get rewarded if the terms of the deal are good. Plus it gives ‘big tech’ the mechanism to make more deals faster at a time when deals are scarce.
Also the cofounders often make a decent return out of the deal and have a strong incentive structure to continue making their technology successful inside the new, better resourced parent firm.
All-in-all, it is pretty good for everyone.
True you don’t make the higher multiples that you might have made had you waited… but you make a decent return, take some risk out of portfolio and get to declare a victory.
Maybe this structure should be used outside of AI more often?
This has me thinking whether this can be applied more often to other non-AI tech categories more frequently.
Because the structure makes a lot of sense.
Sell the access to your core technology to a deep pocketed parent firm and put some cash into everyone’s pockets, while still operating the company and growing it in niches that do not compete with the parent firm.
From a founder’s perspective it’s almost like the best of both worlds… sell your technology and lock in your exit. Then you will need to help the acquirer but still get to continue to work on your company as a smaller, independent player.
