Practical Artificial Intelligence for Ecommerce  

Everybody is talking about artificial intelligence (AI). But is it real, or just the buzzword of the moment? More importantly, can ecommerce managers actually put it to use today?

Let’s start with the basics: what is AI? AI is a broad category of technologies within the field of computer science that seeks to train machines to act intelligently. In other words, data scientists work to train machines to plan, learn, reason, and solve problems, as well take on a wide variety of tasks traditionally done by humans (e.g. self-driving cars, build Ikea furniture).

Although there are many ways to achieve artificial intelligence, machine learning is particularly prevalent in the ecommerce and digital marketing and advertising advertising spaces. Machine learning is an approach to AI in which data is used to train machines. In a classic case, data scientists feed a machine the right answer to a question, and the machine learns to find right answers on its own.

For instance, let’s say you want to train a machine to find consumers who are new to your brand and have the potential to become high-value customers. The machine can look at your existing high-value customers to in order to assess relevant attributes and online behaviors that are common among them. Once it identifies patterns of behaviors that indicate someone has a propensity to purchase your high-value products, you can then target them with ads to bring them to your site.

Although AI sounds as if it’s the stuff of science fiction, it’s actually widely used today, in a wide variety of use cases:

  • Programmatic Advertising: Most digital ad campaigns are executed programmatically, meaning media is purchased on an impression-by-impression basis, based on the consumer who will see the ads. Most programmatic platforms leverage algorithms that are trained to learn which consumers, in which channels, are most likely to respond to ad ad.
  • Customer Profile: Big data analysis of customer purchase behavior is used to proactively identify important milestones (as Target famously did).
  • Product Merchandising: Content-filtering and collaborative filtering recommendations Engine can help ecommerce sites determine which products consumers are likely to be interested in, even if they’re first-time visitors to a site.
  • Pricing Strategies: AI can help identify the optimal selling point on a product-by-product basis in competitive marketplaces, such as Amazon. Algorithms examine a host of attributes, including popularity of a product, current competition, profitability and so on, to recommend a price to offer a customer.
  • 1:1 Messages and Website Experiences: Algorithms update product recommendations and website pages based on what real-time browsing data says about customer interest. For instance, if a website visitor is looking at a fancy red dress, the recommendations engine may recommend appropriate red shoes to match.
  • Front-Line Sales Support: Chatbots are becoming increasingly common. Chatbots are software that can simulate human-like conversations, allowing website visitors to ask questions like, “where’s my order” and receive a plain English response, such as “Let me check on that. Can I have your order number?” By analyzing questions and topics, chatbots have the ability to learn, which means the more they interact with website visitors, the better they become at helping them.

 

Benefits of AI

Although many people fear for a variety of reasons, it’s really quite helpful to ecommerce sites. To begin, it can do things that are impossible for a human to do. It can crunch through massive datasets to make smart decisions in real time and look at numerous attributes that affect the cost of a product, and recommend the optimal cost to offer a consumer in order to win the business AND drive profitability.

It can also plough through data to find insights that are otherwise unknowable. For instance, Dstillery, an AI company, used unsupervised machine learning to help a yogurt manufacturer discover a lucrative, but surprising, customer base: junk food addicts! This is critical: marketers assume they know their audience, AI forces us to test those assumptions.

AI also adds incredible efficiency to operations, as evidenced in numerous use cases, from focusing ad spend on the consumers most likely to respond to an ad, to chatbots providing instant customer service to any website visitor who has a question.

Best of all: you don’t need to fill your ranks with mathematicians, data and computer scientists in order to reap the benefits of AI. Most technology companies have built AI directly into their platforms and services, which means you can leverage it in your merchandising, pricing, personalization and other efforts today.