AccuWeather, a global leader in weather information and digital media, recently launched its AccuWeather bot for Messenger, or Abbi, an AI-powered weather assistant that handles plain-language questions and provides weather information and localized forecasts on Messenger. The bot helps people access accurate and personalized weather forecasting information and uses conversational English to answer questions about forecasts worldwide or other weather-related questions.
We recently caught up with the team at AccuWeather to learn more.
In 100 words or less describe what your bot does?
AccuWeather for Messenger, or Abbi, is a personal AI-powered weather assistant. It helps users quickly find the current conditions and forecasts for their exact location, and millions of other locations worldwide. Abbi provides exclusive and proprietary weather forecasts such as RealFeel® Temperature, and AccuWeather’s MinuteCast®, a minute-by-minute precipitation forecast for the next two hours. Ask Abbi if an umbrella, sunglasses, or a jacket is needed today, and it will make appropriate recommendations based on your forecast. Abbi is available to help users make more informed decisions throughout their day, all with Superior Accuracy™ from AccuWeather.
Who are you trying to reach?
The AccuWeather bot for Messenger is for any user looking for accurate, personalized, weather forecasting information to help them better plan their day. Abbi provides a one-of-a-kind experience, conversing with users in natural language (English) about forecasts worldwide, or any additional weather-related questions they may ask her.
How did you go about designing your bot?
We started with the basics. On a daily basis, people worldwide send AccuWeather ZIP codes, city names, and country names, in search of accurate weather forecasts. Before Abbi, people were using Messenger to ask for personalized forecasts. Once we were confident we were answering most frequently asked questions, we began focusing on the conversations people were engaging in with Abbi. We wanted the experience to feel natural, and the responses to show it understood what users were asking in order for it to be of most use to them. We also provided users with options they could select from, while they were conversing with Abbi, allowing them to find the weather information they were searching for more quickly. By utilizing all the tools available through Messenger, such as quick replies, templates, and an open chat box, we encourage users to interact with Abbi in comfortable and familiar ways.
What have you learned since launching? Any surprises or successes you want to share?
Having a reliable process to get enough information to provide the user with accurate, up-to-date forecasts and weather information is essential to Abbi’s functions. However, forcing users through an onboarding process can be unappealing. Our first modifications to Abbi were aimed at creating a better initial experience for our new users. We learned that those who received a successful response to their first question will continue to engage with her.
What are you going to build next?
Part of AccuWeather’s mission is to personalize weather data so people can make better informed decisions and improve their lives. Over time, users will find Abbi personalizing their experience even more. As Abbi learns, it will be able to cover topics that are important to our users, giving them the ability to plan events and activities with confidence.
Facebook is a place where people come together to discuss the weather, share their weather-related experiences and content, and stay informed of upcoming weather forecasts and conditions. Having Abbi available 24/7 through Messenger gives people another way to access personalized weather information quickly and easily.
Our Building Bots for Messenger series highlights different experiences on the Messenger platform. These businesses and developers have approached building their bots in a unique and interesting way, and are seeing success on the Messenger platform. Look for more bot profiles right here on the Messenger blog.