How machine learning is shaking up e-commerce and customer engagement

Profile picture for user barb.mosher By Barb Mosher Zinck April 5, 2017
Approaching e-commerce from a narrow transactional view doesn't work anymore. The e-commerce wins go to companies with better experiences and fewer shopping cart abandons. That's where machine learning can help - but how? Barb Mosher Zinck explores how Sitecore uses machine learning to change e-commerce, as well as digital marketing and content analysis.

Now that machine learning has moved out of the hands of a select few and is accessible to a broader market and set of use cases; it’s good to look at how different technology providers are leveraging it to improve and expand their products in the name of a better customer experience.

How can you take advantage of machine learning and AI to improve your e-commerce experiences? Ryan R Donovan, SVP of Product Management at Sitecore, took me through some tactics Sitecore implements in its CMS and e-commerce solution.

There are two key focuses for Sitecore’s use of machine learning and AI:

  • Efficiency in operations: How can you streamline the content authoring process?
  • Better insights to optimize: Are you achieving your desired outcomes?

Adding value through machine learning

Donovan pointed out that Sitecore is focused on providing machine learning-based features that provide real value (as opposed to a great demo). Machine learning (ML) supports digital marketing and improved content development (which improves engagement) in a few ways:

From a content perspective, it performs semantic analysis to:

  • Auto generate taxonomies and tagging
  • Help improve the tone of your content by analyzing for things like wordiness, slang, and other grammar-like faux pax

From a digital marketing perspective, ML can:

  • Help detect segments of your customers or audience
  • Improve the effectiveness of your testing and optimization processes
  • Provide content and product recommendations that increase the engagement time a customer spends on your website.

And from a backend perspective, it can help with fraud detection, something that every company with an e-commerce model needs to monitor actively.

One example where Sitecore uses machine learning is through Path Analyzer, a tool that analyzes clickstreams to see how well your contacts are interacting with content and how that matches your defined business outcomes. Path Analyzer first came out a few years ago, but Donovan noted that since the release of its second generation in August (as part of Sitecore 8.2), it has evolved from a cool demo to something that delivers real value.

Path Analyzer tells you things such as what content is working and what’s not, where the points of abandonment are and which tests you conduct are the most successful. All this information helps you improve your content, as well as the flow of content and product information on the website to support a specific business objective, such as selling a product.

Open connectivity and a clear customer data strategy are key

Sitecore may be developing its own machine learning capabilities to do things like Path Analyzer, but it’s also making major investments in its product roadmap to support open connectivity with other platforms. The big one it’s working on now is building in support for Microsoft Azure machine learning. But Donovan pointed out that Sitecore wants to enable truly open connectivity to other best of breed platforms (e.g., Amazon ML).

Donovan spoke about the need to support open standards for data connectivity, to create what Sitecore refers to as “immersive engagement.” Immersive engagement is delivering the right experience in the right context across channels. That cross-channel experience and the sharing of context means you have to pass customer data back and forth between the technology products that support each channel; the key being bi-directional data exchange.

To do bi-directional data exchange, you need to have a clear customer data strategy. Matching your customer profile properly across channels is just the start. You then need to track how a customer is engaging across the channels and whether that engagement is in the context of a sole purpose, or different purposes. You also need to know where and what they’ve done prior and if that relates to what they are doing now. It’s a complex process that a human simply can’t do, especially not at scale.

The evolution of e-commerce is not just about selling products

E-commerce is obviously about selling products, but it’s not all about selling products. That’s why it’s not as simple as turning on an e-commerce shopping cart, popping up a list of products and waiting for the money to roll in. There’s an entire experience you must create now that involves the right content, the right products, all channels in the purchase journey and understanding the shopper intimately.

You can’t wrap your head around all that without leveraging machine learning in some way.

Donovan said the machine learning is a side effect of the convergence of digital strategies and touch points. So if you want to understand how a customer is moving across touchpoints as they move through a purchase processes, and if you want to incorporate content that will increase conversion opportunities by examining what the customer has looked at in the past, what their interests are, what products they already have that are complimentary (and this list can go on and on), you’ll need to adopt tools that leverage ML.

He also said the algorithms are getting smarter and if the data is there, connected properly, businesses can take advantage of ML to enrich the data and extract the best value and insights from it to drive businesses forward.

My take

Many e-commerce experiences have improved greatly, but many more are still caught in the traditional approach of product lists and shopping carts. Probably the best example of where companies are leveraging ML for e-commerce is with remarketing across search and social media.

But there’s a lot more these companies can do with the web/mobile experience to improve conversions. Tools such as Path Analyzer from Sitecore can help because they force you to look at the entire journey (or the path) people take as they move through a shopping experience and they give you the insights needed to take product research to a higher level. Other e-commerce providers (at least ones that come from similar vendors) offer tools that help analyze the journey and build the experience beyond the traditional shopping cart as well. Now we just need more companies to use them.