We asked you to respond to the opening rounds of our Big Conversation and what follows is an attempt to distill an important issue which lies underneath many of the responses. By I.D. Scales.
We're talking commoditisation. There's a well-grounded fear of the invevitable commoditisation of major parts of the emerging M2M IoT ecosystem.
At least it's a fear expressed by some players. For others of course, commodity is what they're banking on. Or at least, they see a market about to be driven by volume and billions of connections and in the IT business that usually means a lowering of costs, prices and generally less revenue per user. They're building that scenario into their plans.
Give us your thoughts and further questions on these issues, either below or email them to me at firstname.lastname@example.org (some people like to keep their communications private for obvious reasons).
Lowering costs and prices in one part of the ecosystem, however, usually creates opportunities for the rest, so there's a tension there.
ARM, for instance, can see a sustainable and important position for itself in the emerging ecosystem as volumes become large enough for it to justify the expense of designing and supporting sensor silicon. It needs the rest of the market to gel to build that volume and it's prepared to go to some extra trouble to help ensure that it does (see - The Big Conversation: Too many M2M standards or just not the right ones?
But commoditisation is more problematic for some players. Mobile operators, for instance, fully understand that M2M ARPU is not going to be anything like what they're used to with good old-fashioned human subscribers, but they are nevertheless nervous of the connectivity price pressure that will result from a horizontally segmented market. They would, ideally, prefer to offer (or work with partners who offer) end-to-end M2M solutions.
Cosm is a platform play for Internet of Things developers and to me its most pertinent aspect is its business model which is a variation of the so-called 'Freemium' model now well understood as perhaps the dominant monetisation route for many online businesses.
Here's the thing about so-called OTT, online businesses (and why 'fremium' figures).
As there is no floor to the marginal cost of the software/service part of the equation (the cost of storing and distributing service just keeps on lowering with no end in sight), there is no pricing model that is 'safe' for the long term - it you do price your service there is always the chance (nay, the certainty) that another player will come in and undercut you.
Let's step back a little.
The imperative for any packaged or "off-the-shelf" software or systems business is to "write once, sell many times". This has been understood for a couple of decades now. Much of the cost here is "up-front" in software development. The idea, then, is to amortise the bulk of the software engineering costs over multiple users. The more potential users you think you can snare, the lower you can set the pricing, the more likely you are to win customers and so it goes round in a virtuous circle (or vicious cycle if you're a competitor with a different business model).
With Web services this dynamic steps up a gear. While off-the-shelf software may offer incremental revenues when customers want the software 'tailored', each sale also incurs significant customer acquisition and support costs. In effect, while the economics of write once sell many are clearly very powerful, there is still a significant 'marginal cost' for acquiring each extra customer.
Web services, however, tend to squeeze those marginal costs even lower, based as they are on commodity hardware (cloud-like services), ever cheaper Internet connections and, very importantly, open source software and platforms. When the resulting services are offered (at least initially) for free, then those pesky customer acquisition costs are squeezed further, while support can often be charged out as a revenue earner (instead of an overhead).
According to Usman Haque of Cosm, its open platform monetisation (for at some point someone has to pay something) comes essentially when something physical needs to happen. There are revenues associated with deploying or managing sensors, for instance, which can be channeled back to fund the platform.
There is a kind of irony here: as with Skype, for instance, which essentially makes most its money from arbitraging switched calls (not VoIP calls or video), so too with Cosm which envisages making money when a man and a screwdriver is involved in what is otherwise a fully Web service sort of activity.
This is clearly a different game from the one M2M has traditionally played. is it one that we may be seeing a lot in the emerging Internet of Things realm and, even, in corporate M2M applications?
Let us know what you think.
please sign in to rate this article