Chatbots save lives

Together, we are in the midst of a public health crisis and all we can do is wait and stay home. We’re slowing down the rate of infection, saving time, giving services time to prepare, but while the world waits for a cure, time is our treatment.

The world is moving online now more than ever. The services we use every day, government, health, finance, insurance, and communications are trying to cope with this change. Customer enquiry surges are reaching critical heights, overwhelming customer support, leading to dangerous wait times.

So chatbots save time and chatbots save lives.

Saving time and lives

Our lives are still going on at home, using the same services we were using before isolation.

We’re waiting longer for our services and sometimes are unable to even use them.

Just this week I’ve had several bad experiences. My internet provider told me to not call unless it’s an emergency. My internet is down and you are an internet provider… No one is calling you for an ambulance.

I was also shopping and Kmart’s website made me queue for their website, waiting 30 minutes in a virtual lobby to buy some socks.

Other websites have just been unable to provide answers or services completely. 

And we’ve seen the results:

But hold the phone, or hold the hold music.

I’m not saying we replace customer service with virtual chatbots.

There are real issues with virtual customer service and we know them too well. 

Press 1 to speak to a human, we tell our entire issue to a phone line. This call is then transferred to a customer service department that hands you to someone else who asks the same questions you answered at step 1.  

Our craft as writers, and my job as a conversational designer (chatbot builder), helps ease these unpleasant moments we all hate.

“How can I help?” is the obvious question, and we need to work with organisations and people to find the answers.

How can chatbots save lives?

New information becomes available every single day which makes it hard for everyone to keep up. We surge websites, hotlines, and customer support teams. Millions of people make enquiries sparked by the same events and sources. 

Companies are letting customers down, governments are unable to give people answers.

If you design a conversation around a service or set of answers, it can answer 80% of customer queries. But automation doesn’t replace staff, it supports them, by answering the questions that can be answered without a human. This frees up vital resources to answer the important questions only a human can.

I’ve done customer service in a call-centre myself. The feeling of reading from a script someone else wrote, unable to help a customer who asks a question you can’t answer cause it’s ‘off script’ still makes me cringe.

I’ve even been asked if I was a robot.

Customer service jobs will become more empowered with automation and encourage more individual agency for staff to focus on complex or unique queries.

How much life does a chatbot save?

Let’s see how conversational automation can save customers, company, and staff time.

Australia’s government welfare service, Centrelink, has recently been inundated with COVID-19 related applications for the same benefit, all being asked the same eligibility questions. 

In 2020, the average wait time for Centrelink’s customer service team is 15 minutes and 44 seconds. In March 2020, there were 250,000 enquiries made in one day. Automating some questions or answers for customers’ questions would save 15 minutes on every single call.

So let’s ask the bot:

The funny thing is, creating a bot that could respond efficiently to 50 specific and frequent customer questions only takes 72 hours to build.

The art of listening well

Another way chatbots save lives is through automation. Conversational automation does more than just answer questions. Nowadays with natural language understanding in AI, we can design for listening.

AI can be taught to register keywords, sentiment, intents while a customer is speaking to a customer service staff member. This information can then be used to feed into services, transcripts, or data.

This example is a prototype from McDonald’s America. (It’s cool, but I wish I had examples of better companies doing it.) It’s live transcribing an order through drive-through and ordering into McDonalds’ Ordering Structure.

With AI listening to conversations, customer service agents can really listen to their customer rather than recording.

This removes the dangerous possibility of human error, especially in a crisis where misinformation could mean negative financial or health implications, not just missing out on sweet n sour sauce.

Although, we are human. We aren’t perfect. Since conversational automation and bots are designed and built by humans, human error will alway