Harnessing AI in Health Care Marketing: From CRM Data to Implementation Solutions

Although evolving privacy laws, HIPAA updates and the phasing out of certain digital tracking (i.e., tracking pixels) are designed to protect consumer privacy, these changes present significant challenges for health care marketers striving to meet patient and consumer expectations.

The use of artificial intelligence in health care marketing has emerged as a powerful solution. It can analyze vast amounts of data while ensuring compliance with privacy regulations, personalize digital interactions with patients and predict consumer health–related behavior.

In a presentation during SHSMD’s 2024 AI Seminar, Leveraging AI in Healthcare Marketing and Communications, Kylie Worley, a marketing automation supervisor at Ochsner Health, and Christina Paik, Ochsner Health’s senior marketing specialist of CRM, emphasized AI’s varied roles in health care marketing. They addressed general questions about AI advancements, discussed specific opportunities in health care marketing for predictive and generative AI, and explored tactics that health care marketers and organizations can employ to ensure that AI is used properly and strategically.

Predictive AI in Health Care

Predictive AI has been a valuable tool in customer relationship management (CRM) for several years. It allows health care marketers to identify patterns in past events and make predictions about future behavior. Patients can be segmented based on demographics, patient history and care needs, helping to identify health trends and patients at risk for future conditions.

Worley explains that Ochsner Health uses predictive AI to create patient propensity models that foresee the likelihood of patients needing specific services and identify those at risk for certain diseases, such as hypertension or cardiac events (Figure). These models help in the development of targeted campaigns, including social media targeting. The models also allow Ochsner Health to more effectively utilize its marketing budget.

“Using the propensities within our program, we’re able to enroll patients into a specific [marketing] campaign,” Worley says. “We’re able to add dynamic content to existing campaigns to create more of that personalized experience. We’re also able to create targeted Facebook lists, … [for example,] targeting patients who would need orthopedic services or sports medicine services.”

Generative AI in Health Care

Generative AI, on the other hand, requires a human touch to be successful, as it creates new content from prompts and involves combining and refining AI-generated content.

The history of generative AI shows its evolution from simple algorithms to advanced tools that can perform research, generate content and provide personalized responses.

To effectively use generative AI in marketing, it’s crucial to maintain a consistent brand tone and establish AI personas to build trust and engage with the audience.

Prompt writing is essential, as the quality of AI-generated content depends on the specificity of the prompts. It’s also important to review and edit AI-generated content.

When selecting AI tools, “consider your audience,” Paik notes. “Consider if the tools that you are planning to use have the correct features that you need … to engage with that audience and really help you reach [your] marketing goals. Take a look at … user reviews, the ease of implementation and the training needed.”

There are also risks for health care marketers to consider with generative AI: the loss of a competitive edge as AI tools become more widely accessible; increased consumer expectations for personalization; the high initial costs of specific generative AI tools; and the need for fact-checking AI-generated content due to possible inaccuracies and bias.

Preparing Your CRM, Protecting Patient Data

Health care marketers can optimize AI’s benefits by preparing their CRM and understanding the best applications for their teams.

“Prepping your CRM is an essential first step to keep your patient data safe and to avoid any legal issues down the line,” Paik explains. She advises health systems to follow these steps:

1. Perform an initial data audit. Know which data are used in your prompts and avoid using any sensitive customer data. Remove any personal identifiable information from your AI tools.
2. Set up some privacy protections, allow your patients to opt in and out, and set usage policies on how your customer data are going to be handled.
3. Secure your data access. Determine which authorized users have access to customer data, and continually monitor and audit that access on a need-to-know basis.
4. Test your processes and mitigate any potential issues by checking your AI content for bias error. [If you encounter unintended results,] make sure that you’re giving your team members a hands-on experience in training that they need to utilize their tools to the fullest.
5. Have a data backup plan. Enable quick recovery for restoring your data in case there are any data loss or errors and to ensure data integrity in order to comply with any legal requirements.
6. Continually monitor these processes. Set up security policies that can detect any potential threats to your customer data, and then continually review and audit that sensitive data for the future.

Getting Started

Ochsner Health discovered tools within its CRM that its team was previously unaware of and sought a road map of new AI opportunities from its CRM provider.

Of note, many vendors are willing to integrate AI with their CRM. Even without a CRM, health care marketers can still leverage numerous free or third-party AI tools—such as ChatGPT, Jasper AI and Canva Magic Write AI tool—by researching their features and trying out free versions or demos.

“A great place to start [if you don’t have a CRM] is with generative AI,” Worley says. “Within our marketing department, we have people in various roles using it for all sorts of different things, … whether it's optimizing a webpage or social [media post]. The really nice thing about using generative AI from that perspective is being able to take one piece of content and repurposing it for multiple platforms without having to rewrite it for the multiple platforms.”

Worley also emphasizes the importance of conducting an initial skill analysis for the team to understand individuals’ interests and strengths with AI, including understanding machine learning basics, identifying and reducing bias, and becoming a generative AI expert. She suggests tailoring professional development accordingly, incorporating free resources and training, or certifications (if there is a budget for such expenditures).

Measuring Success

Measuring the success of AI-driven campaigns involves tracking key performance indicators (e.g., click-through rates, conversion rates), conducting A/B testing and analyzing click activity. Regularly reviewing these metrics helps ensure long-term success and effective use of AI optimizations. Increased user engagement, click-through rates and conversions lead to a higher overall campaign return on investment, demonstrating the economic benefits of implementing AI in marketing strategies.

Balancing Concerns and Benefits in an Increasingly AI World

Now more than ever, the health care sector is exploring how to balance privacy concerns with AI advancements. The three primary concerns for health care marketers regarding AI adoption include the fear of faulty results or inaccurate information, inherent bias in AI tools, and the risk of becoming overly dependent on AI, according to Paik.

The success of any AI program relies on deliberate reporting, A/B testing and understanding what works best for the team. By optimally leveraging AI, health care marketers can deliver more effective, patient-centric marketing strategies while overcoming some of the challenges (e.g., privacy laws, HIPAA updates) inherent to the health care marketing setting.