Over the past year I have worked extensively with a range of “IoT companies” (and I explain those quotes later) that required my knowledge about the Internet of Things including hard- and software development and/or data analytics and machine learning. These interactions came naturally, and my involvement was typically an obvious next step when my partners were lacking that skill set or experienced staff in their teams. While my work and contracts could be described as strictly B2B and I’m glad I’ve never had to do hardcore business development like cold calling potential customers, I was often doing work for particular external clients and projects. These clients didn’t care, in fact, in most cases I think they were happy to have dealings only with one company and not with an entire host of subcontractors. In a way, the companies I worked with could be described as systems integrators, although none of them would actually describe themselves as such.
The more I begin to understand how business works (remember, I have a 50% academic + 50% engineering mindset, plus a big mouth), the more I realise two things:

  • Friend & Foe in business is gone. This is probably pretty obvious in a time when Apple sues Samsung over mobile phone patents, while at the same time Samsung is involved in producing iPhone parts… It is the same in the IoT: While some companies are direct competitors for the parts of the cake, they may be customers of each other for other parts and may even have to collaborate on joint projects. This convolution is getting more complex every day, as most

  • IoT companies are systems integrators. This may be in contrast to the title that speaks of “focus”, but let me explain. Complete IoT solutions that address a particular need often feature bespoke hardware components including embedded code, off-the-shelf communication devices such as gateways, a data platform and backend software running in the cloud, as well as data processing and/or visualisation capabilities. Imagine your company is absolutely brilliant in doing one of those things, but has little expertise in the others? Well, it’s simple. You don’t sell. In fact, most clients getting started in the IoT will ask for a proof-of-concept, and I can assure you there are shrieks of terror when you’re a software startup and need to provide a hardware prototype to demonstrate the capabilities of your algorithms with just two months notice. At that stage, your business model suddenly shifts from selling good software to organising solutions that require you to understand and decide on things which are completely out of your comfort zone.


For that reason, there is pressure on IoT companies to collaborate. I’m not talking about gigantic pick+shovel manufacturers in this gold rush, the likes of Amazon AWS or Microsoft Azure. I’m talking about normal companies trying to pay the bills by selling IoT directly to those who need it, be it hardware, software or services.

As far as I can see it, there are currently three prevalent models to sell IoT solutions to industrial clients:

  • Consultancy services that are totally agnostic to the technology they’re recommending. They explain why IoT is good and make a business case. They make their money based on (hopefully) good technical decisions, and one might think they’re providing objective information, as they’re going to be paid no matter which vendor the client ultimately decides for.

  • Specialised start-ups, which are probably top-of-the-line for their particular niche and that have to stretch a bit to meet the demands of clients. Much like the consultants, they should be providing good information on the bits that complement their products, that make their product look good as part of a complete IoT solution. (This also explains the earlier quotes around “IoT company”: If you’re doing sensors, you’re not automatically doing IoT, but to sell them you may have to become an IoT company)

  • Full-stack providers are companies that offer bespoke solutions for most industries. These are typically larger and established companies that can easily draw on existing internal resources and that have teams of hardware and software developers. From what I’ve seen, many clients would love to work with a full-stack provider, but fear the cost of the big names in the field and therefore favour interactions with smaller companies and start-ups.

I think this latter scenario gives rise to good opportunities. In fact, it’s already happening. Many well-established companies are indeed willing and able to accommodate external solutions if they appear better suited. For example, most players that are involved in the industrial IoT now also offer their own IoT cloud solution, but everyone knows that companies like Bosch and Zühlke offer integration e.g. with the Thingworx IoT Platform or EVRYTHING if it is the right thing to do. On the other side I’ve seen more than one occasion when OpenSensors was approached by customers to provide an all-encompassing solution, but it remained a big question mark whether a bigger partner would even be interested developing some bespoke hardware in collaboration. This is where both sides need to communicate better their willingness to interact.

Collaboration is key to interoperability.
Interoperability is key to collaboration.

I believe that this level of collaboration and interaction is healthy for the development of IoT as a field and a win for everyone involved. The IoT can only live up to its promise if we stop treating the Internet as a replacement for a wire; i.e. if we can provide data integration across devices, companies and entire verticals. A rich ecosystem of data platforms, a choice of different gateways and devices that can speak to each other, and vendors that are not afraid of recommending competitor products in cases where their own product is not the best solution are key to breaking up silo thinking and pushing for interoperability. If we as a field get this right, I think everyone will benefit from the mix and match of IoT products.

Rather than selling only complete end-to-end solutions, the market for mix and match must be infinitely larger.

Not only because of initial sales, but because clients are much more inclined to experiment and replace components of their IoT solution if things “just work”. Thus, we must not only look good to our clients, but also to our competitors that may choose to interact with us in their next project. And, as a start, that “looking good” can be done through adhering to open standards on the technical level, and fairness on a business level.