Listen to an audio recording of this blog:
Some of today’s biggest technology companies are changing the way we find and interact with health information by expanding their use of chatbots. A chatbot, which can be powered by voice or text, is a messaging tool designed to act like a conversation or interaction. Chatbots, also referred to as conversation agents or “bots” for short, are making interacting with technology a lot like talking to a person.
Advancements in technology continue to offer new ways to communicate HIV information. Here are some of our recent findings on the state of bots.
Bots have gained recent momentum for the following reasons:
- The use of messaging platforms has exploded, surpassing even the use of social networks in active users.
- Developments in artificial intelligence, machine learning, and natural language processing - which are the cornerstones of effective bot interactions – enable bots to have increasingly engaging and human conversations.
- Off-the-shelf technologies, like the use of third-party apps on Facebook Messenger, have made it relatively easy and inexpensive to build and test bot prototypes.
We see a lot of potential for using bots for HIV communications. Some (not yet developed) examples include:
- Helping a user find information on a website more quickly
- Answering simple questions about who should get an HIV test and where, when, and how an individual can get tested
- Assisting a patient in scheduling or rescheduling a clinic appointment
- Enhancing an existing mobile application by offering customer service
If you’re exploring or building your own bot, here are some industry best practices:
- Assess your organization’s needs - Review your communication goals and your users’ needs to see if a chatbot is a good fit.
- Build for a realistic and simple use case - For current bots, the best use cases are for simple tasks that can be solved in a short conversation.
- Focus on the content - Many of today’s most effective bots use a clear script, mirror the target audience’s language (including slang words), and use friendly and inclusive language that repeats input from the user to check for understanding.
- Ease users into interacting with the bot - Using familiar elements (such as buttons) and preset responses, gives users clear steps and can make first time users feel more comfortable.
- Set user expectations early - Let users know they are talking to a bot, not a human, and be clear about what the bot is capable of.
- Improve user retention through research and testing - Focus on the user’s initial interactions with the bot, which is where most users drop off, and conduct user testing to improve the experience.
Bots are already making a positive impact on health and health care, including applications that help to provide rapid diagnosis, point to optimum health care solutions, increase awareness of social issues, and motivate the right actions. Here are some additional blog posts:
Watch for additional content from HIV.gov on bots – Don’t forget to sign-up to have our articles delivered to you by email so you don’t miss a post!