What your internet of things startup should learn from Netflix and the disaggregation of TV


The internet, standards and open platforms have allowed businesses to break up content and services into discrete and disaggregated parts. Now instead of paying $80 a month for a pay-TV package and getting 60 channels you don’t want, you can assemble the shows you like via Netflix(s nflx), iTunes(s aapl) or any number of other available services and get a relative bargain.

The problem of course, is that when you break things up, you might get exactly what you want, but you no longer have the assurance of the cable company delivering your single stream. Now you have Netflix’s servers (which are really Amazon’s), a content delivery network or two, an ISP and even your home Wi-Fi network all a factor in the quality of the experience. So when a video fails, who does the consumer call?

netflix google tv second screen 3 This a problem that the video industry is working out right now, with efforts from ISPs, big content companies and even startups trying to ensure quality of service on the network. But our TV isn’t the only place in the living room where this drama will play out, the smart home itself is an example of this same potential problem. And in the case of connected devices and sensors you actually have even more players working together, and no viable economic model just yet.

Stefan Ferber, the VP for business development in the Internet of Things & Services unit at Bosch Software Innovations recognized this in an awesome series of 10 points to consider about the internet of things (read it all). But as a consumer trying to piece together my own smart home (as opposed to going with a closed system) his seventh point resonated — the idea of accountability with the internet of things. From his post:

Our current understanding of accountability traces back to a rather black-and-white picture: when a business contract is not fulfilled or an illegal action is conducted. Our usual reaction is to go back and check what went wrong. But how can we trace the root cause of a defect, e.g. in a line of code, in our super-connected world with billions of devices, billions of users, and millions of decisions taken every second? We can’t just shut down the internet to investigate an incident. With lots of stakeholders from software agents to operators, service providers, system builders, programmers and users, the only way to establish accountability in the IoT is by creating trust-building mechanisms that are not focused on single-company or personal responsibility but on collective accountability. This leads us toward a rather collective insurance model and less to a personal accountability mechanism.

So what will this look like with the internet of things? I think to get to that point you have to figure out all the links in the chain. And unfortunately, we don’t even know that yet. For example, some of my devices are connected and controlled via a SmartThings hub, but others are controlled via an independent app or by If This Then That (IFTTT). To complicate matters more, some of my SmartThings devices talk to others in my home via IFTTT. All of this runs through my router, which I purchased myself and then over my ISP-provided modem out to the public internet. Some of these companies keep my information on their own servers, while others rely on Amazon Web Services.


And those are just the services. The sensors themselves could fail, either because of a battery or even poor quality construction. In the case of hardware it’s much easier to find the person responsible and ask for a fix, but what happens if the maker of the hardware has gone out of business. SmartThings may grow to become a huge and dependable company, but right now it is a startup with $3 million in funding and about 18 months in existence. If the company goes away, what happens to the cloud-based service affiliated with the hardware? WE’ll be discussing elements of this in several panels at our Mobilize event next week, where we focus on building out an ecosystem for the internet of things, from the bandwidth to the business models.

These aren’t easy problems to solve, or even articulate, which is why Ferber’s post got me thinking. I’m eager to hear from you guys how to create an accountable system for the internet of things. Of course, I could always buy something from Comcast(s cmsca) or AT&T(s t) or maybe even the new Staples offering, and get the relief of one responsible party to call when things fail. While the best of those systems will combine that responsibility a willingness to continue innovating with new products and offerings, they will build closed platform that make it tougher for me to leave.

That’s probably how most consumers will find their accountability, while on the corporate side big companies like GE(s ge), Intel(s intc) and others will offer similar packages.  Then the real question will be how those on the consumer side — where benefits aren’t always measured in ROI and margins are slim — share in the profits of the ecosystem. Those fights might be just as brutal as the ones we are watching on the video side or even on the consumer web.


Rob Freeborn

the answer to this problem has absolutely nothing to do with open source vs COTS vs custom SW.

this is an issue based in the ability to deliver root cause analysis at a spectacular scale – understanding the correlation between virtual and physical resources, applications, hardware, core/premise and third party services.

It’s the service provider that can make all those behave as a “service” as opposed to discrete elements that will deliver value to the consumer.

The IoE will only go as far as it’s ability to avoid the “fail whale” scenario.

Tim Enwall

In some sense, companies like Apple and Microsoft faced this issue with operating systems (and even today with the app store). Software APIs were part of the accountability infrastructure between fairly complex systems and 3rd parties. The 3rd party used the operating system API and counted on that API being tested and remaining stable — which it did sometimes and didn’t other times.

I worked at Apple for 8 years in the early days of the Mac OS. The accountability ran deeper — there were numerous app bugs that were fixed in the OS — and visa versa. The culture of Apple and their 3rd Party developers focused on the customer experience, which I think improved that consumer accountability.

I doubt that’s the whole answer, but software APIs that are open, documented and stable can certainly play a big role. The more proprietary the app or the device, the harder it is to drive system-wide accountability.

Comments are closed.