Déjà vu? Machine Learning’s current growth echoes IoT’S from a few years earlier, but without the fizzle

Posted in Rubikloud Perspectives
By Aaron Dauphinee on July 12, 2017

Recently Entrepreneur published an interesting article on how small businesses with Internet of Things (IoT) products and services can avoid being locked out of this lucrative market by larger companies.

Guest writer, Maciej Kranz, vice president of corporate strategic innovations at Cisco, advocated in his article that start-ups need to “throw assumptions and principles that worked in the previous technology waves out the window” to gain an advantage. His view is driven by observations in how the IoT market quickly accelerated a few years ago after IoT was placed on top of Gartner’s hype cycle in back to back years combined with the acquisition of Nest by Google.

Kranz shares that he experienced an onslaught of undifferentiated companies all making proprietary devices that connect various sensors and who concertedly tried to appeal to multiple industrial segments. Very few had vertical expertise with deep understanding of their customer’s industry, needs, and business challenges, which is why they struggled. He also advocates not going it alone and recommends partnerships as another means to further yourself with selling to a specific sector.

His ultimate recommendation for IoT small businesses is to “step outside of the traditional role of a technology company that sells to IT, service providers, or consumers. The goal of IoT is to address the specific care-abouts of [line of business (LoB)] decision makers in specific vertical markets. More and more, LoB executives are the ones making technology purchase decisions.”

RK’s Perspective:

Hindsight is always twenty-twenty, and it’s easy to provide this advice after seeing that old means of selling can no longer be applied to a new technology. However, the parallels between the market acceleration for IoT then and what, we believe, will be a similar pace for machine learning (ML) in the coming 12-18 months, are uncanny.

While Gartner currently doesn’t expect the mainstream adoption of machine learning for two to five years, given recent market dynamics we suspect they may soon early up their view. Forrester is already more bullish than them on insights-driven businesses that will fuel growth by utilizing their data and adopting ML and artificial intelligence.

Two recent “Google Nest-like” catalysts fall within the retail sector, which we see as the hotbed for quickening ML’s adoption into products that drive tangible benefit to consumers. The well-reported shift into offline retail by Amazon with their purchase of Whole Foods is one; giving them a premium playground to further strengthen their existing ML capabilities. But it should not be overlooked that Wal-Mart continues its drive into acquiring niche online retailers, such as Jet.com, Modcloth, Moosejaw, and Hayneedle, as they pick up Bonobos for $310 million, which introduces the everyday low price retailer to new customer segments.

Both add competitive intensity, with the blurring of offline and online retail plus expansion beyond core customers, in a sector seeing iconic brands shutting their doors and that is primed for ML adoption to drive automation and efficiencies. More importantly, this need for change in a specific sector is what will drive ML retail products to deliver real value to consumers and, ultimately, avoid the subsequent disappointment from undifferentiated IoT products with no real viable consumer utility.

Is this all déjà vu exactly? Perhaps not. But it is clear where the momentum is headed on a need for retailers to change to remain competitive and we expect that the adoption of ML solutions in retail will push it into the mainstream.