The Power of AI to Deliver Personalization

When personalizing at scale, tools like AI are the only way to go.

Anyone familiar with artificial intelligence knows that Alan Turing is often noted as the ‘father of computer science and artificial intelligence, despite that the idea of artificially intelligent beings and forms goes as far back as Greek and Roman mythology. While the first instance of AI being recognized as a genuine field of science did not occur until the Dartmouth Conference in 1956, interest in the field has ebbed and flowed throughout history. AI shifting from research concept to engineering application in 2015 was monumental, but it is still in infancy. We are living through the ascent of AI, with no telling where or what the pinnacle will be.  

However, in its growing buzz, it’s easy to feel ‘already behind’ when powerhouses like Apple, Facebook, Microsoft, and AWS are rolling out ground-breaking changes to how we interact with machine-learning technology around the clock. In actuality, there is no catching up when everyone is a competitor. It’s how you utilize and implement all of these new technologies to fit your business, big or small.

Artificial Intelligence is an umbrella term that includes multiple technologies, such as, machine learning, computer vision, natural language processing (NLP), and data mining, among other emerging ‘AI buzzwords.’ These technologies, in essence, are performing analysis and modeling from a multitude of different data sources and trying to provide an output that mimics the way an intelligent being would respond to stimuli and context clues in the same situation.

AI Landscape

Stated in an Adobe AI Blog, “By 2020, businesses that use AI and related technologies like machine learning and deep learning to uncover new business insights will take $1.2 trillion each year from competitors that don’t employ these technologies.”

The influence of AI is already extending into many aspects of modern society, regardless of overall complexity. From the most technically savvy to the average consumer, AI is affecting the way in which we continually operate our day-to-day lives.   These influences for the average internet user can be seen when navigating to a merchant’s website and having a chat bot ask about your customer experience or exemplified in targeted ads that begin appearing as you skim your preferred social media site. The browsing habits we form, the tweets we post, the articles we read, and the sites we shop all leave a paper trail of data for retailers to analyze and then act upon. In many ways, the data we create by using the web is more important than us, as physical consumers.

As data collection advances, experience optimization will occur at a 1:1 level and eventually be predictive. The vision for technology of the future are systems that provide the most direct path to your answer, that already know whether you will take a left or a right.

There are certainly a number of growing business use cases around artificial intelligence, but an example of a company hitting it out of the ‘AI ballpark’ is Stitch Fix.  Stitch Fix is an online styling service that delivers a personalized shopping experience right to your doorstep. Each shipment contains five items of clothing, shoes and/or accessories for you to try on at home, keep what you like and return (prepaid) what you do not. But how is personalized styling on the cutting edge of AI?

Stitch Fix states that they are “weaving data science pervasively throughout [their] business,” and in fact, they’re doing it really well. Through proprietary algorithms developed by different ‘Algorithms teams’ broken down by operational stage, Stitch Fix is changing the way men and women shop online.

Their teams are:

  • Client Algorithms – Understand our clients, predict their demands, and optimize operations
  • Styling Algorithms – Meet the recommendation engineers: where the art and fashion is quantified and shipped
  • Merch Algorithms – Mastering the science of clothing, inventory, and logistics
  • Data Platform – Build robust and powerful technologies to store, access, and compute data

“We use trained neural networks to derive vector descriptions of 'pinned' images, and then compute a cosine similarity between these vectors and pre-computed vectors for each item in our inventory.” (Stitch Fix Algorthims Tour)

Natural Language Processing (NLP) is used to ‘score’ items based on the client's request note and textual feedback. The implementation of NLP allows Stitch Fix to harness large amounts of explicit data, both from clients' self-descriptions and from clothing attributes; all providing immense accuracy if algorithms are employed to consider that data.

Natural Language Processing from Stitch Fix
Natural language processing from user entered text by customers. 

By looking at their company needs and assessing ways that AI can expedite their processes, Stitch Fix has been able to streamline online shopping even further than before; all while making things even easier for the consumer. Now, that’s machine learning at work. AI and ML are not all or nothing. Just because you can't create the next bot doesn't mean there aren't incremental ways to bring your business into the fold of AI. Maybe it’s not about staying cutting edge but applying these new features into businesses in real time. Whether that’s going full-bore like Stitch Fix and winning big, or picking small ways, with data, to get on the AI train.