We continue our series on market themes we like, a combination of bottoms up observations and best practice from portfolio companies and the broader ecosystem with top down research. At Episode 1 we still believe the best days of tech investing are ahead of us – the scope for building $1b businesses has never been greater.
Theme 5: Smarter Infrastructure – So, a Robot Walks into a Bar in Shoreditch…
This one might take a while – the connection of absolutely everything to the Internet and what we find interesting about it is a long topic with a complex setup. It’s a bit like one of those incredibly corny jokes you hear in a bar, which after about ten minutes relies on some terrible pun for the payoff. Well, the payoff in terms of what we like in this space comes next week, but first you need to stick with us through the preamble. So imagine we are in a hipster bar somewhere in East London, open a craft beer or pour yourself a Spirulina-matcha smoothie and sit comfortably, this may take a while. I can’t even promise it is going to be very funny.
Let’s begin at the beginning with some discussion of terms. IoT may take the award for being the tech meme to move most rapidly from great industrial hope to bottom of the hype curve. Not long ago 10’s of billions of things were going to connect to each other to create trillions of dollars of new economic value – and it was going to be happening right about now if I remember. Sadly, as economists like to say, you cant really see it in the numbers yet. In my view, this is unfortunate, as it has led many to underestimate the impact that all the technologies and business models coalescing to make this happen will have. Skeptical journalists like nothing better than to dismiss IoT by trivializing it as nothing more than connected fridges automatically reordering milk for you. Which aside from anything else is the wrong problem – trendy techie types only drink almond or oat milk. No one drinks actual dairy anymore – as if!
“IoT may take the award for being the tech meme to move most rapidly from great industrial hope to bottom of the hype curve.”
So what went wrong? Well, there’s lot of reasons this hasn’t happened as fast as people predicted (aside of course from the obvious one that things seldom do, which is why hype cycles exist). Many early use cases were trivial, software was glitchy, connectivity and sensors didn’t interoperate properly, there are too many competing standards – and for many B2B applications security, data protection and operational performance just did not meet spec.
“Many early use cases were trivial, software was glitchy, connectivity and sensors didn’t interoperate properly, there are too many competing standards.”
But this disguises the real progress that has been made and the ultimate impact it is going to have – it’s useful to review some of the major developments that have taken place and that may have slipped peoples attention. Firstly, many manufacturers have quietly added an awful lot of IoT tech to their portfolios and Installed Bases. Automotives, industrials and tech companies have all invested heavily in putting sensors and connectivity into their equipment to make it “smart”. They’ve also built lots of software to manage it and in many case moved some ways up the path to thinking about themselves as service providers or software companies – deliverers of outcomes rather than vendors of hardware.
Secondly, there has been considerable interest in the development of horizontal management platforms. Let me clarify. Where installed bases are fragmented rather than concentrated in one manufacturer – e.g connected home, connected office or the shop floor – the most likely player to provide a management platform is a seemingly neutral third party. Which is why you’ve seen acquisitions like Cisco buying Jasper, PTC buying Thingworx, Nokia and Withings, Microsoft with Solair to position Azure as a IoT platform – and corresponding moves by AWS and IBM pushing their cloud IoT platforms. Lots of lessons have been learned, scale has been created and new capabilities have become available to the market.
Thirdly, the underlying technology is working its way along the compelling logic of Moores law where as scale goes up, performance increases and price comes down. Whether you want low power consumption, radio enabled, temperature resistance or video capability, Systems on a Chip (SoC) are getting more powerful, cheaper and specialized. Similar trends are taking place in MEMs and sensors – helped in no small part by capacity and capability built up by the mobile phone supply chain. Again, many acquisitions in the semi-conductor space reflect this – Cypress buying Broadcom’s IoT business, Sony buying Altair, Intel buying Yogitech and of course ARM moving from acquirer to being acquired in the groundbreaking Softbank acquisition.
“the underlying technology is working its way along the compelling logic of Moores law where as scale goes up, performance increases and price comes down.”
Fourthly, there has been huge progress made on standards, with industrial groups coming together at the device, connectivity or application layer to agree standards. Big platform brands like Apple and Google have created SDKs for health and home with varying degrees of openness. And finally, there is tons of innovation at the data science, data visualization, sensor, edge, security, telemetry and connectivity layers that provide a wealth of enabling and supporting capability.
So the net of all this is IoT applications are no longer limited by the basic technologies. This is such a huge, interconnected space that an enormous amount of organization and setup work needed to be done in order to create an environment and ecosystem within which creative entrepreneurs can innovate. Because in the end this was less about building another set of applications on the Internet than it was about building a new Internet – and although by no means finished, that work is well underway. At some point soon we will reach the tipping point and all sorts of new businesses will start to appear. Which is why any day soon two young men with beards and a collection of pre World War 1 aircraft insignia tattoos on their arms are going to come to you and pitch their idea for a chain of 7×24 fully automated vegan-porridge vending machines, with a customized seed topping based upon a DNA sample. Are we in the future yet?
“Because in the end this was less about building another set of applications on the Internet than it was about building a new Internet.”