ways to use machine learning

Machine learning has been described as the science of getting computers to act without being specifically programmed to act. While for some, that definition can bring to mind nightmarish images of machines taking over the world, it is actually a great way to automate marketing activities to make them less time consuming and to provide better experiences for your customers.

But how do you get started with machine learning? In this blog post, I’ll share the two ways you should be thinking about it this year: machine learning for your customers, and machine learning for you (the marketer). Let’s dive right in.

Machine learning for your customers

First, marketers can use machine learning to power online customer experiences. When many think of website personalization, rule-based comes to mind first. Rule-based personalization refers to the ability to manually set up business rules to deliver specific experiences to different segments of people. For example, you could use rules to ensure that only visitors within the US see references to free US shipping throughout your site.

However, you can also use machine-learning algorithms as a more scalable way to achieve unique experiences for individuals (i.e. 1:1 personalization), rather than segments of people. You are probably most familiar with this type of personalization in the form of recommendations for products or content. But machine-learning personalization can also be leveraged to recommend other aspects of your website, such as categories, subcategories, brands, promotions and more. You could also use them to dynamically modify site navigation, search results and list sorting.

ways to use machine learning
List sorted for a particular shopper using machine-learning algorithms

Essentially, every aspect of your website can be driven by machine-learning algorithms. How does it work? Every time a visitor engages with your site, you learn more about him. You learn the categories and brands he engages with most. You learn his favorite colors and his preferred price point. You learn his favorite blog topics or authors. Machine-learning algorithms leverage all of this information to select the right experiences and recommended items for each individual. And by showing him the most relevant content across your site, you can help him more easily find what he’s looking for, leading to more conversions and improved loyalty.

Machine learning for you (the marketer)

Beyond using machine learning to fuel the experience for your customers, you can also use it to help you focus your attention on the highest priorities for your business. Marketers have so much data available to them from many different sources (often only accessible by different members of the team). It’s impossible to stay on top of all this data at all times, and it’s not always easy to prioritize the biggest opportunities or the biggest threats.

Machine learning can be used to cut through the noise. It can make sense of all the signals in the data to help you identify patterns, opportunities, or problems based on your key business metrics, and alert you to a shift so you can respond quickly.

Use machine learning to analyze which of your campaigns is providing the highest business impact, to recognize where opportunities exist on your site to help you plan future campaigns, or even to identify when a problem arises with your existing campaigns or general site performance. For example, machine learning and predictive analytics can analyze your typical inventory levels, taking into consideration seasonality, day of week, and general variation, and recognize when visitors are seeing more out of stock items than usual. This allows you to quickly identify the problem and take action immediately.

Machine learning identifies that out of stock items are higher than predicted

Final Thoughts

Machine learning is quickly becoming a hot topic in the marketing industry, but many marketers are still working to figure out the best way to leverage it in their own strategies. As you put the finishing touches on your 2017 plan, think about machine learning on two fronts: how it can impact your customer experience and how it can help you better do your job.