It's been 20 years since the phrase "Internet of Things" was coined -- yet the industry hasn't quite lived up to its prophesied potential.
IoT darling Nest, acquired by Google for $3.2 billion in 2014, was destined to be the trailblazer for the industry.
Valuations for IoT startups all around the world literally hinged on the vision that Nest's acquisition sold to the market - to design intelligently automated experiences with data collected from devices.
Unfortunately, that vision to bring the holy grail of big data to IoT has largely remained a promise.
In fact, with the trajectory that companies in the industry are currently taking, big data will never be the future of IoT because of one big problem: companies who don't know what to build also don't know what to track. And because they can't learn from the data they don't track, they continue not knowing what to build.
Companies who don't know what to build also don't know what to track. And because they can't learn from the data they don't track, they continue not knowing what to build.
This catch 22 is further amplified by the 90% problem - where the embarrassingly manual data munging of any organization can take up to 90% of engineering resources, impeding not only insight but also impact.
Additionally, new formats of data from heterogeneous devices not only make old questions hard to revisit, but also make new questions tediously difficult to answer without manual work.
Iterating on data becomes just as slow (if not slower) than iterating hardware.
In summary, big data remains glaringly absent in the foreseeable future of the Internet of Things.
Without actionable data, IoT companies are light years away from intelligent automation and even more, product market fit.
And the "killer app" that every IoT company dreams of achieving? Just as absent as the data.