The 2 Main Ways to Use Machine Learning in Your Marketing

June 26, 2017 Katie Sweet

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.

Many marketers in smaller teams think that machine learning isn’t for them — that it is best used by the major names like Netflix or Amazon. But that doesn’t have to be the case. Machine learning can be used by companies and marketing teams of all sizes. In this blog post, I’ll share the two main ways you should be thinking about machine learning for your marketing: machine learning for your digital experiences, and machine learning for you (the marketer). Let’s dive right in.

1. Machine learning for your digital experiences

First, marketers can use machine learning to power digital experiences. When many think of website personalization, a rule-based approach 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, or that only visitors from a certain industry are invited to join a webinar.

In these situations, rule-based personalization lets you target a specific message or experience to a group of people (i.e. segments) that fit a few specific criteria. But it is not ideal for one-to-one communication. If you want to create individualized experiences based on the preferences of each individual, you would have to set up and manage hundreds or even thousands of rules. That is just not scalable.

Machine-learning algorithms represent a more scalable way to achieve unique experiences for individuals (i.e. 1:1 personalization), rather than segments of people. You are probably already 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.

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.

E-Commerce Example

A shoe retailer could choose to show trending products on its homepage, recommending shoes that are most popular at a given time. To personalize those recommendations, it could boost the brands and price points each visitor prefers. One visitor may see primarily Vans shoes while another may see Steve Madden — based on which brands each visitor has shopped on the site. This helps shoppers quickly and easily find new shoes they are more likely to be interested in, rather than show all new shoes to them whether they are interested or not.

B2B Content Example

Now let’s explore a content example. Assume a person landed for the first time on a site for team productivity solutions. She navigates to the resources section of the website and begins to search for a general term related to the space. In the search results, the site could prioritize resources related to her industry, product interest, challenge and her stage of the journey to surface results that are more likely to be relevant to her. So even though her search is general, the search results can help her find appropriate resources more quickly.

2. Machine learning for you (the marketer)

Beyond using machine learning to fuel the experience for your visitors and 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.

E-Commerce Example

A retailer could use machine learning and predictive analytics to analyze typical inventory levels, taking into consideration seasonality, day of week, and general variation. Machine learning can recognize when shoppers are seeing more out of stock items than is expected, helping the retailer to quickly identify the problem and take action immediately.

B2B Content Example

A site focused on demand generation could use machine learning to analyze content downloads on the site, to identify when the number of leads generated is lower than predicted. The marketing team could dig into that information to identify when an important link is broken to correct it before too much damage occurs.

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 explore machine learning, think about it on two fronts:

  1. How it can impact your digital experience, providing highly personalized, relevant content to customers when they need it;
  2. How it can help you better do your job by identifying your best opportunities and helping you to take quickly advantage of them.

 

About the Author

Katie Sweet

Katie Sweet is Content Marketing Manager at Evergage, focused on crafting relevant content and maintaining the Evergage blog. When Katie isn’t meeting challenging deadlines, chasing blog contributors, and driving content at Evergage, she is constantly reading, homebrewing, traveling, and trying adventurous foods.

More Content by Katie Sweet
Previous Video
Super Personalized Human Marketing—At Scale
Super Personalized Human Marketing—At Scale

www.hartehanks.com/human Customer engagement just got bigger and better. Harte Hanks and Wipro have partne...

Next Article
The Difference Between Integration and Orchestration is Customer Context
The Difference Between Integration and Orchestration is Customer Context

To keep up with shifting consumer expectations and deliver a context-driven customer experience, you need t...

Don't Miss a Thing.
We'll keep you up to date with our latest content you'll love.

Stay Updated