Artificial intelligence (AI) is captivating industries everywhere. Today, you can find AI powering voice assistants, chatbots, and even generating content. Now, businesses and marketers are looking to utilize AI to automate tedious tasks and enhance the customer experience. Due to this, MarTech suites are responding in kind.
Earlier this year, Salesforce unveiled a multitude of generative AI suites including Marketing GPT, Commerce GPT, Sales GPT, Service GPT, and more. With many new AI tools being introduced that marketers can utilize to enhance the customer experience, lets discuss how AI can help you add a further degree of personalization to that experience and how you can ensure any dynamic content sent to customers is accurate and relevant.
Generative AI is excellent for brainstorming and generating dynamic email content, especially if you are on a small marketing team. Writing multiple variations of email and ad content can be incredibly time-consuming and monotonous, so it’s natural to think the time could be better used focusing on objectives that mature your marketing operations.
One of our Level Up cohosts, Laura Madden, Sr Manager of Lev’s Marketing Strategy and Services team, recently reflected on this in a recent episode, “In my early marketing days, I remember writing SEO copy for a promotional products company. They had about 30 different types of promotional pens, and I was running out of clever ways to say, ‘You should put your company name on this pen and sell it.’ It was a hot mess, so I’m thinking ChatGPT could have been very helpful to me in avoiding burn out.”
Platforms like ChatGPT, and Salesforce’s upcoming Marketing GPT, make it easier to generate alternative versions of the same copy, so you can A/B test what’s most effective with your audience. To test this, our team did an experiment. We prompted the following to ChatGPT: “Provide five email subject line options for a retail company providing outdoor gear to people with differing levels of expertise around the outdoors, ranging from hobbyists to professional outdoor athletes. The user persona I’d like to target is a hobbyist outdoor camper between the ages of 20 and 32 years old, female, owns a dog and goes camping one to two times a month. The email content is intended to provide personalized gear recommendations based on recent purchases this type of customers have made before and should be no more than seven words.”
Getting this specific, ChatGPT gave us a few great subjects:
Being able to generate these subjects on-the-fly is incredible as it can empower you to truly focus on improving the customer experience.
The age-old question so many email marketers have been faced with is “What's the best time to send my email?” If you serve multiple personas and segments, there’s so many factors that determine what is the best time. As marketers, we’ve all been in situations where we’re scratching our heads on why an email performed better at various times throughout a week. Fortunately, AI doesn’t just help with generating content, but also figuring out the best time to deliver that content.
Tools, such as Einstein in Salesforce Marketing Cloud, offer send time optimization (STO) to help you determine the best time to send your email content to different segments. STO analyzes your customers’ activity, such as when they are opening an email and clicking links within it, and then completely automates the send time of your emails.
You still have a further degree of personalization when it comes to sending emails by providing STOs windows. Perhaps you want to send a specific email seven days later, or sometime in the morning to ensure your customers have time to shop during the day? You can provide STO these specific guidelines and it will choose the best time to send the emails in the window you specifically provide. This way, you can ensure that your excellent messaging and promotions are sent at the right place and time.
With today’s AI-powered tools, it’s even possible to implement dynamic web content. Salesforce’s Marketing Cloud Personalization (formerly known as Interaction Studio) captures behavioral triggers and then can display custom content based on those triggers. For example, if a customer views, adds to cart, or purchases a product, those behaviors can be captured to build dynamic, relevant product recommendations.
Another use case is displaying custom call-to-actions. For example, let's say a prospect signed up for a demo of your product. You could capture that event as a behavioral trigger and then display other helpful products and marketing materials that supplement their upcoming demo instead of delivering call-to-actions that continue to hound them to sign up for the demo. But AI is not something you can just set and forget. It takes iteration and training to ensure AI provides the results you want.
A misconception we see with AI is that some marketers think that AI will do everything for them. Unfortunately, AI is only as good as the data you give it. It’s common for large language models, like ChatGPT, to confabulate facts—a phenomenon commonly known as hallucinations. When you ask it to provide you content, it gives you what its programming thinks you want to hear. It does not have the capability of human thought to determine whether something is factual or not. You must fact check the content generative AI programs give you as a result.
Prompting itself is an art. How you ask generative AI programs what you want is going to affect the result it gives. If you are unsure how ChatGPT comes up with a recommendation or generates content, ask it how it came to that answer. It will provide you with its reasoning on why it provided the content it did, and you can refine your prompt from there.
When experimenting with new AI-powered tools, curiosity is your best friend. Be curious, experimental, and skeptical. It’s easy to get sucked into the popularity and just trust whatever content it gives you. Just have metered expectations and continue to refine your prompting skills. The more AI learns from your prompts, the better and more accurate it becomes.
Looking for more AI best practices you can incorporate into your marketing operations? Check out our other AI marketing related content.