One of the open secrets in retail is that conversion rates haven’t changed much in years. And while retailers have devoted a lot of time, energy, and money to the problem, we’re still stuck at about 3% conversions. Those recurring messages from retailers led me to hop in a car and travel from Silicon Valley to San Fran to meet with Sentient Technologies. The goal: to learn whether the buzzword of AI that has taken Silicon Valley by storm is an empty promise or if it has real-world applications in sectors such as online retail.
I sat down with three key people in the Sentient family for a deep dive into “deep learning”:
– Babak Hodjat, Co-founder and CEO of Sentient Technologies
– Todd Hay, Product Owner for Sentient Aware.
– Jon Epstein, Senior VP of International at Sentient Technologies
In a nutshell, here are the two ways Sentient is commercializing AI as it pertains to our online seller client-base:
Aware uses deep learning to understand each individual customer’s preferences as they shop, letting you understand their sense of style to personalize product recommendations.
Ascend leverages genetic algorithms and marketers’ inherent creativity to optimize websites from thousands of possible designs.
Now, you don’t have an AI company without a very heady visionary leader, in this case, Babak Hodjat. He has been in the AI field for decades and has the patents to prove it. And you don’t have a visionary leader with only two ideas and only focused on commercialization.
So, apart from Aware and Ascend, Sentient has a host of other AI units and initiatives. They work with big institutions like MIT and Oxford on projects that leverage AI to do things like predict the onset of sepsis in the ICU or maximize crop yield in controlled growing environments.
Sentient also has a robust AI research arm, with 12 awarded patents in a variety of AI disciplines from deep learning to evolving neural networks. Sentient also has a fully-autonomous AI hedge fund and a distributed computer network that ties together 5,000 machines to tackle vast AI problems.
They also temporarily embedded an AI chip into my brain so that I could understand most of what Babak was talking about.
Here is what happens when I ask Babak if we are going to achieve Artificial Intelligence in our lifetime.
Bringing things back down to commercialization, here are the videos profiling the two main commercial products they are delivering for online sellers.
Sentient Ascend is an AI-powered conversion rate optimization (CRO) solution. If you know anything about A/B testing, it’s similar to that, but instead of trying a couple ideas at once and seeing which performs better, it enables companies to try as many design ideas as they want simultaneously and find the ones that convert best. Those ideas can be pictures, copy, layouts, colors, or pretty much any change a brand might want to test. And since it uses a branch of AI called genetic algorithms and a different branch of statistics than traditional tools, it can search through tens of thousands of possible combinations and find winning performances in the time it would take to run a single A/B test.
Sentient Aware, on the other hand, personalizes shopping experiences for each customer through the power of artificial intelligence. Most sites use cohort analysis to “personalize” their offerings, but really, what they’re doing is selling to “users like you.” With Aware, it’s different. The AI analyzes each image at a granular level, understanding product characteristics like color, fabric, or shape, and subtler characteristics like the relative height of a high heel or the scoop of a neckline. Each customer click trains the AI, teaching it what that individual user is interested in, so that it can provide more and more personalized products as the shopper moves through their buying journey.
What the full extended interview series created and posted through Prodygia here:
For more great content about the different technologies that are transforming the way companies do business in the eCommerce world, subscribe to our YouTube channel, or check out our partner sites!
Article originally published by Josh Halpern at our partner organization, Getting to Global.