Why Mobile App and IoT Companies Need to Get AI into Their Products Now
Personalization: not just a “nice-to-have.”
Last week, Amazon CEO Jeff Bezos shared with an audience in Washington, DC, about the pivotal role machine learning has played for the $430 billion company. While the firm’s AI pursuits have spawned intriguing devices like Alexa, a voice-activated assistant, Bezos insisted the real power of Amazon’s machine learning engine lies in things like improved “search results” and “product recommendations for customers.”
The real power of Amazon’s machine learning engine lies in things like improved “search results” and “product recommendations for customers,” according to Amazon CEO Jeff Bezos.
If you’ve browsed Amazon.com, then you’ve seen the razor-sharp precision of its product suggestions and the magic of its personalizing algorithms.
Much like its fellow emperors of the Internet (Google and Facebook), Amazon has built its dynasty, in part, by offering highly personalized experiences to its users. The company achieves this by feeding machine learning models reams of user data that the models then convert to behavioral profiles. These profiles then become the blueprint for personalizing features that draw users in like rip tides.
Amazon built its dynasty, in part, by offering highly personalized experiences to its users.
There are certainly factors at play here that I’m not privy to. But in any case, the approach I’ve described has worked quite well for the Seattle-based company. Just look at its sway in the marketplace. It puts long-time incumbents on their heels, creates and dominates entire markets, and beats earnings predictions to boot. Not to mention, the e-commerce giant was just voted #1 in the American Customer Satisfaction Index, an honor given based on a poll of 10,000 people.
I should clarify that this post is not about Amazon. It’s about the preeminence of personalization.
As noted earlier, algorithm-led personalization is firmly entrenched on the Internet. But it has yet to take hold in the realm of mobile apps and the Internet of Things.
To be fair, app and IoT device makers are limited by the narrow data they collect about users. Amazon, on the other hand, can know a great deal about a person due to the eclectic nature of the products and services it sells. I can only imagine the detailed profile Amazon has on me based on the wares I seek out and purchase on its website.
At any rate, the hurdle of a thin data set should not deter companies from personalizing their products. For sure, organizations that treat AI-driven personalization as a mere “nice-to-have” will put themselves at a disadvantage. Given enough time, that disadvantage could spiral into a deathblow.
Companies that treat AI-driven personalization as a mere “nice-to-have” will put themselves at a disadvantage. Given enough time, that disadvantage could spiral into a deathblow.
A reasonable objection gets raised when one considers building a machine learning solution for their app or IoT device: it’s hard and expensive.
For starters, it requires data scientists and developers with machine learning expertise — highly coveted talent — to create the models. Then, a steady stream of clean, rich data is needed for training the algorithms. Add to that months — if not years — of grappling with the models until they’re reliable while introducing new ones to the mix, and you’ve got yourself a costly little enterprise, indeed.
Thankfully, there’s a remedy for this: AI as a Service.
The technology sector’s renewed interest in artificial intelligence has touched off something of an AI “renaissance,” to borrow a term from Jeff Bezos. Market research firm IDC predicts the cognitive systems market, currently valued at $8 billion, will soar to beyond $47 billion by 2020.
What I’m driving at is, given the spirit of innovation around AI, odds are a group of smart folks will invent the kind of algorithmic engine that app and IoT device makers need to personalize their solutions. App makers can then simply purchase the machine learning service as a subscription. Indeed, one startup already has launched such a service.
Early adopters are sure to jump on personal AI services as they begin to emerge. When they pass this on to their customers in the form of rich personalized experiences, these trailblazers will gain a strong foothold in their industry. Whether that foothold will rival Amazon’s is not for me to say.
About the author: I’m the content writer at Neura, Inc, a startup whose personal AI service drives engagement for mobile apps and IoT devices.
