Artificial Intelligence

How AI M&A is going to rise

This is a contributed piece by Julie Langley, partner at Results International

It won’t surprise anyone to know that M&A activity in the Artificial Intelligence space is on the rise.  But the level of upturn in the last couple of years is staggering. In research we’ve recently conducted on the sector, we identified 98 M&A transactions globally in the AI sector for the period January–October 2017, up from 74 and 44 in the whole of 2016 and 2015 respectively.

Our analysis revealed a number of other interesting insights:

The buyer universe is expanding. In the period 2011-2015, there were 13 companies which made two or more acquisitions in the AI sector, primarily the large US tech giants such as Google, Apple, Facebook and IBM. In the period 2016-2017, the equivalent figure has risen to 20 companies, and whilst the US tech giants still dominate (with Apple overtaking Google as the most acquisitive), the range of acquirers has expanded. Companies as diverse as Ford, GE, TomTom and Axon are stepping up their activity. The imperative to future-proof traditional industries will continue to drive new buyers into the sector.

Core AI is no longer dominating M&A activity.  Whilst “Core AI” technology, such as autonomous systems, driverless cars and computer vision, attracts most of the headlines the reality is this only represents about a third of all M&A activity in 2017. AI businesses being acquired have developed narrow domain-specific applications which solve a particular problem for a vertical industry, such as healthcare or finance, or a function, such as marketing or HR. These applications aren’t always very newsworthy – improving pipe-laying or preparing insurance quotes – but they solve real problems for the buyer and increase revenue or reduce costs.  Which leads into my next point.

Valuations are more rational than might appear.  As many of the companies being acquired are pre-revenue, there is sometimes a tendency to view them purely as acqui-hires and to view value on a “price per employee” basis. Hence Magic Pony selling into Twitter for a reported $150 million becomes “$10 million per employee”. In practice in most deals valuation is driven not just by the quality and size of the team, but by the target having developed a machine-learning based solution which fixes a problem or opens up a new revenue stream for the buyer. Take the example of Magic Pony and Twitter. A key part of Twitter’s strategy is live video streaming. Ensuring high quality video however is challenging given the bandwidth available. Magic Pony had developed a patented technology which enhances image quality in video. There is a business plan behind these deals which goes far beyond the value of the team. Patents also play an important part in many high value AI deals.

AI skills are in demand regardless of location. Whilst the US continues to dominate, buyers are increasingly looking outside the US for talent and technology. 61 of the 98 AI deals announced in 2017 involved US sellers, equal to 62%. This is down from 82% in 2015. Buyers are increasingly open to acquiring AI development teams across the globe and to leave them in situ once acquired. This has not always been the case. Several (although certainly not all) of the large US tech players have in the past tended to want to keep their development teams on the West Coast and acquired talent has been asked to relocate. With the skills shortage in AI, it seems there is a willingness to accept distributed development teams, in exchange for retaining the best teams.

The question often gets raised as to what room there is for startups in AI and machine learning in an area increasingly dominated by the tech giants.  Notwithstanding the clear battle for talent, the data seems to suggest there is a lot more room for startups to grow and thrive than might be expected.  Rather than take on the tech giants directly in core or generalized AI, very often they are in fact using the AI frameworks offered by Google, Amazon, IBM and Microsoft to build very specific AI tools at the application layer to solve problems for specific industries or functions.

2017 felt like a tipping point in the adoption of AI and ML, and M&A activity has followed suit. AI is on everyone’s agenda, and as businesses seek to future proof their offer through the application of AI, the acquisition of relevant solutions will be high on their list of priorities.


« Why you need a bootcamp grad


Lessons from CA's internal startup accelerator »
IDG Connect

IDG Connect tackles the tech stories that matter to you

  • Mail


Do you think your smartphone is making you a workaholic?