This week I’m at the Things Conference in Amsterdam, an event focused on the IoT low power radio access network technology LoRa. I’ve been speaking with dozens of companies and the key commonalities between them are a strong heritage in their space (be it street lighting or animal monitoring or whatever) often acquired over decades and that they provide a more-or-less end-to-end service, comprising hardware, connectivity, software platform, integration and consulting.
This week I also spoke to a mid-sized vendor I’ve known for years. They shared that they’ve been winning supplementary business in a non-core area by virtue of understanding the clients’ business and having a hard-earned relationship with that client. There are better products out there that fit the client requirements, but the deep knowledge of the customer segment and the simpler integration are enough to persuade the clients to work with them.
There’s a common theme emerging for me: that of expertise and understanding of a particular vertical trumping technical capability. Tales abound of vendors pivoting products to address the next new shiny thing with little but an aspirational powerpoint behind it, while truly capable vendors with decades of heritage seemingly struggle to realise (in both senses) the opportunity they have in front of them. What we’re seeing now in IoT is that finally the players with the expertise in an area are getting the upper hand.
The question I ask myself is: are we in a post-platform world? Are the idiosyncrasies of each client so pronounced that every IoT implementation needs to be so customised as to render the horizontal platforms almost worthless? Or at the very least to trigger a bifurcation. On the one hand we have the general IT Services Vendors/Systems Integrators/Consultancies which make use of generic platforms to underpin expensive projects. On the other we have smaller specialist vendors supporting their product offering with (usually massively under-priced) consulting.
OK, the reality is much less stark than that. Platforms still offer a tremendous benefit of simplicity and scalability for most IoT solutions. And some functionality is all about scale so is not replicable by smaller vendors. There’s certainly still a home for Azure, AWS etc. But, vendors ignore at their peril the need for deep understanding of the client segment they are pursuing. I’ve talked previously about Libelium’s move to pursue the services space. This is reflecting the reality that long standing product vendors are usually pretty good at helping their clients also to implement them.
I know that anyone from a software start-up will claim that providing services is a distraction from their goal of becoming the next unicorn. It is. But, as I’ve blogged elsewhere, they must not let pursuit of that prize blind them to commercial realities. Or to the fact that often what the client is paying for is not the hardware or the software, it’s the flairware.
It doesn’t matter where you get the flair form. It can be bought, borrowed or built, but it can’t be faked.
This applies to IoT, but it’s also relevant for a wide range of other technology areas that we cover in our DX12, that are less mature. In AI, for instance, I was speaking with H2O.ai last week. While obviously they can apply outstanding ML capabilities to a client’s requirements (Kaggle Grandmasters etc.) it’s impressive how they don’t lose sight of the fact that their clients needs are very specific, and might even have their own secret-sauce. It would be easy to say “we’re the ML experts”.