Posted by Teri Sun

Shared resources almost always come before individual resources. 

Trains and streetcars came before individual automobiles. 

Telegraph offices and then community “party lines” came before personal phones. 

Heck, even bath houses, fountains, and wells came before private, in-home plumbing.

And there's a good reason for that. As technology is being developed, a shared resource is configured to serve many people well enough. But that doesn’t mean it’s ready to be built on an individual basis (having a telegraph office or street car at your house wouldn’t provide the best utility). 

The same holds true for AI chatbots. 

Right now we’re in the shared resource phase with ChatGPT, Copilot, Claude, and others. But that doesn’t mean you should rush to create or integrate a chatbot into your own website. 

These new AI technologies are incredible because they’re built to serve the entire internet. 

But a chatbot on your website should be built to serve your target audience and represent your brand. The magic for your website visitors won’t be in the technology (they’re getting used to that) it will be in how the experience is fully personalized to them and your brand. It’s obvious when it isn’t done right, most AI chatbots feel very transactional in nature because they aren’t trained on the core values that matter to you and your website visitors. 

The standard chatbots may direct a hospital website visitor to a doctor’s profile page when they’re hoping to schedule an appointment. But a well-trained chatbot can recommend specific doctors based on the visitor’s on-page activity, or ask questions that will help them suggest the best option - all in a specific brand voice.

Getting to the typical chatbot experience by implementing a chat interface is easy. The hard part of getting to a well-trained bot is in clarifying your brand values and messaging. Those inputs have a much larger impact on the value of the chatbot than the technology itself. It’s worth taking the time to make sure you get it right and train your bot with the best technology from the start. 

How to Clarify Your Brand Values and Messaging to Train an AI Chatbot

 

Identify Core Brand Values and Tone of Voice

- Ask yourself what emotions or impressions you want the chatbot to convey, e.g., friendly, professional, or playful? List out adjectives that describe your brand and keep them visible as you work on training the chatbot.

- Note specific words or phrases that resonate with your target audience or align with your specific focus (e.g., pediatric care might emphasize clarity and patience).

Outline Key Needs and Pain Points

- Build a list of common questions or challenges that customers bring up in other channels, like email support or customer service calls.

- Prioritize the types of questions that the bot should be able to answer, focusing on those that impact conversion or retention most directly

Develop an Information Hierarchy

- Decide what information is essential for the chatbot to access immediately. Should it focus on facility FAQs, organization policies, health education? An organized hierarchy helps avoid information overload and keeps responses on-topic and useful.

Create Specific Training Data

- Just as with general-purpose AI, the input for your chatbot matters significantly. Write sample responses, categorize common inquiries, and refine prompt styles that best reflect your brand voice.

- Incorporate real-world interactions (anonymized) if possible to enrich the bot’s understanding of common questions and nuances.

Test and Refine

- Implement beta testing with trusted users or team members to gather feedback and catch issues in messaging or response clarity.

- Regularly update your chatbot as your brand evolves or as customer feedback shifts, ensuring it stays relevant and valuable over time.

As you’ve seen from shared resources like ChatGPT, proper prompts and information make the difference between a quality response and word soup. The same logic applies to your own chatbot - garbage in, garbage out. Take the time to build out your AI chatbot and personalization plan so you can maximize their impact for your visitors and organization. 

Topics: Technology, Strategy, Lead Nurture, Best Practices, Digital Marketing, Healthcare, Lead Generation, CMS, Content Management System, healthcare revenue generation