Snow, ice, a hurricane warning, a water main break: the call gets made before sunrise. A superintendent checks the roads, talks to the bus company, and decides by 5:30. Within minutes, the district sends a text, an email, and a phone call, and posts to social media.
Then families start checking. Some missed the text. Some got it and want to be sure. Nearly all of them check on a phone. Pew Research Center finds that 91 percent of U.S. adults own a smartphone, and for about one in six, it’s their only internet connection at home. Many of those phones land on the district website.
What families find there decides how the morning goes. If the homepage still looks like yesterday, they keep searching. If they find last year’s delayed-opening schedule, they act on it. And if the website has a chatbot that answers from the site’s everyday content, “What time does school start?” can get an answer that was right yesterday and is wrong today.
On a storm morning, a chatbot needs to do two things. When the district has posted an alert, it should answer from the alert. When it hasn’t, it should say it doesn’t know and point families to where closings are announced. A confident “yes, school is open” is the worst answer it can give, which is why what an assistant says when it doesn’t know matters most on mornings like this.
Before the first storm, settle three things. Who posts the alert to the website, and does it go up before the text or at the same moment? What exact wording will you use for a closing, a delayed opening, and an early dismissal? And once it’s posted, who checks that the website and the chatbot say the same thing as the text?
Then test it on an ordinary day. Ask your chatbot, “Is there school tomorrow?” If it answers yes without knowing, fix that before the weather does.
With AskK12, one tap pushes a banner alert to every visitor, and the assistant answers storm questions from it.