4 times I got it wrong about IoT and data

Spending the last few weeks sick in Dallas left me with plenty of opportunity for introspection (and little energy for reporting), so I thought a lot about all of the times I was wrong when predicting the future of the IoT. I decided that it might be helpful to share those with y’all since I think it’s always worth exploring my mistakes and trying to figure out why things didn’t come to pass as I thought they would have.

So here are four of my predictions or beliefs about IoT that were just wrong — some of which I’m still really sad about not coming to pass.

1. Data marketplaces: Man, I am still so bummed this one never happened. I really believed that because so many devices and organizations were becoming capable of collecting data — and because it seemed wasteful to have several sensors from different organizations collecting the same data — that we’d establish marketplaces to buy and sell data. Let’s say you wanted air quality data for a specific block in San Francisco, for example; you’d be able to buy that information from a government or even a private citizen who had a specific air quality sensor. Or perhaps you wanted to get building occupancy data so you could build an app telling people when the best time to visit the DMV was; you’d be be able to buy that from various companies.

But rather than a single (or multiple) data marketplace(s) where anyone can buy and sell data, instead we mostly have companies that try to get their sensors in as many places as possible so they can sell access to that data. In the meantime, we still have multiple sensors measuring the same data for multiple companies. So, while we do have data marketplaces, they aren’t as egalitarian as I had hoped. As you’ll soon see, this is a common theme in my missed predictions.

2. Horizontal IoT platforms: Before I covered the IoT, I wrote about enterprise computing for more than a decade. Generally speaking, I focused on the move from siloed enterprise software systems and hardware to cloud computing platforms and deep software integrations that could span industries. I covered a lot of the consolidation across industries like health care, insurance, and manufacturing into one of a few giant software companies that purported to offer “end-to-end” solutions. So when I saw the move toward digital transformation with broad-based sensor, data, and connectivity platforms to help connect different industries, I thought it simply made sense to avoid recreating industry-specific silos that would eventually lead to messy consolidation.

I was wrong. The horizontal platforms weren’t specific enough and many of the companies that wanted to buy IoT solutions didn’t have the technical expertise to add on the layers of customization they wanted or needed. And honestly, after talking to people buying IoT platforms and hearing about their use cases, the related regulatory issues, and their technical skills, I can’t believe I held on to this one for so long. One final issue keeping this prediction down was the fact that a mess of standards and kludgy integrations made even attempting true horizontal platforms that “anyone could program” impossible.

3. An infinitely customizable smart home with sensors: Related to my challenges predicting the future of enterprise IoT, I also whiffed it all the way back in 2013, for similar reasons, when I envisioned the future of the smart home. I thought we’d have a world of sensors that would share their state back to some kind of home hub that would also be capable of controlling other gadgets. Thanks to standards, I was envisioning these devices would be more like commodity hardware but with different design features, and that the real magic would be the interfaces and software with which we programmed these devices and sensors. I thought, in other words, the future looked like a long-forgotten hub from a company called Sense.

The hub was dubbed Mother and it came with generic multipurpose sensors called Cookies. Users would add the Cookies to their hub, then download an algorithm for a use case such as sleep monitoring or motion tracking. In my predicted future, this was how we’d onboard sensors. And from there we would be able to program the hub to use the information from the sensors to make another device do something (think IFTTT). So if my motion-tracking sensor sent motion, a light might pop on. Or if the motion-sensing sensor didn’t track motion after a set time, maybe the light would turn on. This last use case would work for tracking whether or not someone took their medicine or fed their dog.

Instead, everyone built their own versions of light bulbs, sensors, and hubs and created ecosystems around software platforms so only certain devices worked with others in that ecosystem. The Matter standard will help with this, but I underestimated how challenging the security issues would be, the desire of companies to create walled gardens, and the fact that most consumers were not as excited to tinker with smart home gear as I was.

4. Data contracts: This last one is a bit of an outlier, but I truly thought that as companies added sensing and AI to their operations, the data that fed those algorithms would become essential not just internally for the operations themselves, but also externally to others in the supply chain. Thus, I figured that companies would want to share internal data with their partners. But because that data could lead to the exposure of trade secrets or other proprietary practices, companies would want to get the equivalent of NDAs as part of their data-sharing agreements.

This hasn’t really become a thing, to the best of my knowledge, likely because most companies just aren’t ready to share this level of data with other companies. Not because there isn’t a need, but because most digital transformations are extremely challenging to implement, which means companies have enough to worry about with getting their operations digitized. They aren’t ready to try to move their efforts out to their partners.

I’m sure I’ve been wrong in many other ways, but these are the big ones that both stick with me and make me sigh over the missed opportunities they represent. I’d love to hear your missed predictions if you’d like to share.