Skip to main content
Three images of different technology and AI

Automated Marketing and AI: How Smart Tech Can Level-up Your Email Campaigns

By Prasad Saripalli

If you grew up watching the Jetsons, then you were familiar with AI (short for artificial intelligence) before it was a thing. While machine maids and butlers with human-like intelligence still seem like something from a sci-fi movie, AI has become a part of everyday life in many ways. It’s that friendly website chatbot that asks if you need assistance. It helps Amazon recommend items tailored to your taste. It filters spam emails. And it allows Google auto-fill to anticipate your search intent before you finish typing.

The truth is, AI has become an integral part of our daily experience. While we go about our business, AI quietly works in the background to make our lives easier. And, yes, it can make running your wellness business easier, too. Here’s how.

AI gives us valuable insights in real-time

Imagine walking into a casino with 200 chips. Naturally, you want to maximize your return on investment. But you have no clue which slot machines give you the highest chance of winning. Your best bet is to start by exploring different machines, then exploiting the ones where you have the best luck.

This exploration-exploitation trade-off can be summarized by the “multi-armed bandit” problem, originating from a hypothetical experiment. In the experiment, a person must choose between multiple actions with an unknown payout (such as slot machines, known as “one-armed bandits” for stealing people’s money). The goal is to determine the best, most profitable outcome.

In terms of online marketing campaigns, a multi-armed bandit solution is like an advanced form of A/B testing that uses AI to identify which variations of a campaign perform best. Then the algorithm automatically allocates more web traffic to those top performers, putting more revenue in your pocket. Unlike A/B testing, there’s no need to wait for results. AI accounts for results in real-time.

Advanced marketing automation

An email marketing campaign is one real-world example of a multi-armed bandit problem (aka the bandit model). To optimize results, you need to know:

  • The best day and time to email your target audience
  • What types of offers they’re most likely to buy
  • Which subject line will result in the highest open rate

You could spend the time and money to A/B test these things. Or you could use AI to apply the bandit model. With Mindbody’s marketing automation tools, AI does the exploration necessary to uncover the best time of day to email a particular audience segment. This solves for the fact that not all fitness businesses have the resources to perform extensive A/B testing. AI makes it easy, automated, and accurate.

The bandit model can also be used to provide your clients with targeted virtual class recommendations. In this scenario, the algorithm recommends different offerings to consumers on the virtual wellness platform. Based on real-time feedback—such as clicks, bookings, and ratings—the algorithm gains valuable insight into each consumer’s interests. This allows you to provide increasingly personalized offerings.

Increasing impact with AI

Yes, AI can tell you which subject lines, email content, and images perform best. But it can go one step further and generate all those things for you. The result: a delightful email campaign, created by a computer.

In the spirit of continuously evolving, the AI/ML (short for Artificial Intelligence/Machine Learning) team at Mindbody is always integrating the latest technology to improve the customer experience. One AI technology, in particular, known as Natural Language Processing (NLP), helps machines understand and interpret human language. Can you say, “Hey Alexa, remind me to update tomorrow’s class schedule?” That’s NLP at work.

While NLP powers everyday technologies like Alexa and Google, it can also help businesses increase customer satisfaction and revenue. For example, NLP can detect the sentiment of an email. In other words, it can tell if your email contains positive or negative emotions—which might dictate whether that email lands in a spam folder or not.

But the most critical part of any email is the subject line. It’s the first impression that determines whether your email gets opened or not. For example, readers are 26% more likely to open an email with a personalized subject line. And guess what? NLP can read the text inside an email and recommend an ideal subject line.

The AI/ML team recently built a subject line prediction (SLP) feature like this for Marketing Suite, Mindbody’s intelligent marketing platform. Based on historical data from roughly two million emails, this feature automatically recommends impactful subject lines that increase open rates. With Marketing Suite, you don’t need to be an email marketing expert to see great results.

How subject line prediction work

By extracting useful information from subject lines and email content, the AI processing engine groups similar information. In technical terms, this is called “clustering.”

With this method, Marketing Suite’s marketing automation platform defined 16 sets of email templates. Sets were grouped into categories—such as SMS opt-in campaigns and holiday promotions—based on discount percentages, holiday names, and dates.

Once the information is sorted, NLP can “train” the prediction algorithm to rank subject lines based on how well they match different email content. Then, AI technology checks the work to ensure all discounts, dates, and names are correct before returning the best subjection lines to you, the human user.

As great as this sounds, NLP also comes with some challenges. The biggest challenge is teaching machines to understand the nuances of human language. Language is often ambiguous and subjective, which makes it difficult for machines to interpret and automate. It’s also essential to make sure algorithms provide ethical predictions. In other words, we don’t want AI spitting out inappropriate subject lines.

The Mindbody AI/ML team conducted extensive human testing to carefully evaluate subject line predictions based on these challenges. By stepping into the customer’s shoes, we use human judgment to confirm all predictions make sense.

The best part? Based on thorough testing, Mindbody’s subject line prediction outperforms most other solutions—from features to ease of use and accuracy. But don’t just take our word for it. Data projects a 5% increase in email open rates, a 1.5% increase in click rates, a 5% decrease in annual churn rate, and a 33% campaign adoption rate.

The bottom line

While AI will never replace human emotion and empathy, it can help businesses increase the impact of their marketing efforts by connecting with more customers. Subject line prediction plays a role by improving the email marketing experience and boosting open rates—all while saving you time to focus on building real customer relationships.

So what’s next? Mindbody is innovating new ways to improve product features using NLP, such as lead predication and management. With smart technology like NLP, we look forward to new ways to increase conversions and improve the overall customer experience.

See Marketing Suite in action.

Get A Demo

About the author:

Prasad Saripalli headshot

Prasad Saripalli

Vice President, ML & AI, Distinguished Engineer

Mindbody

Prasad Saripalli serves as the Vice President of AIML and Distinguished Engineer at MindBody. Earlier, he served as VP AIML at Edifecs, an industry premier healthcare technology provider, and led the development of Smart Decisions ML & AI Platform with ML  Apps Front. Prior to joining Edifecs, he led similar technology product & platform missions as CTO at Secrata.com, which provides military-grade cloud security solutions, and as CTO & EVP at ClipCard, and as Chief Architect for IBM's SmartCloud Enterprise. He also served as the GPM on Microsoft's Client Virtualization team, shipping Virtual PC and Hyper-V on Windows 7, and as a Dev Manager on the Citrix group that built Citrix Presentation Server (now Citrix XenApp). Prasad obtained doctoral training in Engineering and Computer Science from the University of Florida and post-doctoral training from The Oden Institute for Computational Engineering and Sciences at the University of Texas at Austin.  He taught AI, ML, Advanced ML, Game AI, NLP, Networking, Cloud & Distributed Systems at the Northeastern University and the University of Washington.

New resources, straight to your inbox

Get updates on the latest industry trends, tips, and news.

We're committed to your privacy. Mindbody uses the information you provide to us to contact you about our relevant content, products, and services. You may unsubscribe at any time. View Privacy Policy

Back to top