You have a lot of online messages from customers who want to be taken care of right now. Right. Now.
Does it feel overwhelming? It certainly can be, especially considering how much money is riding on answering those messages.
Hiring people to address those countless messages is one approach, but sooner or later, your employees will have to take breaks. Your customers will be left with their questions unanswered–and possibly even left unread–and they will get ornery.
Rebecca Clyde, cofounder of botco.ai, proposed a solution to this challenge: marketing automation and AI.
But AI in marketing is still a relatively new concept, and companies have been slow to adopt it because of its challenges. The ability to carry on language in a natural way by using software hasn’t been perfected or made available to most of us.
Still, it’s becoming clearer that AI is essential to any successful marketing strategy, and while it’s not at the stage that most people want it to be, it can still prove invaluable not just for your marketing, but for your entire business.
Two billion messages between customers and businesses are exchanged each month on Facebook Messenger alone.
Lots of businesses are failing to respond their messages on Facebook and other social media platforms. They’re probably getting inquiries on a daily basis that they’re either ignoring or simply don’t know what to do with because of the volume.
Some companies just assign multiple people to answer those messages. But business meetings, tight deadlines and vacation days inevitably result in the neglect of some of those messages, resulting in some very upset customers. And a lot of auto-response messages just can’t adequately answer highly specific questions and concerns.
Consumers under 30 have many ways they’d rather interact with your brand and content, most of which have nothing to do with going to your website or the emails you push out. Most millennials (and younger people) are spending their time in messaging apps, not email.
We’re entering an era when the customer of today doesn’t necessarily want to go to our websites, fill out forms and wait for responses. They want instant communication that even the biggest of staffs can’t accommodate.
2. Humans and AI Can Work Together
Humans can make a big difference in the implementation of AI. They can ensure the AI is being trained on accurate, up-to-date content.
In the creation process, ask the bot hard questions and try to stump it. Get creative and bring in a lot of people to make sure it’s as accurate as possible. Think about the real ways in which people ask questions. How humans communicate is ever-evolving, so perfecting the language and capabilities of AI will be an ongoing process that evolves with communication.
3. Examples of chatbots and AI in marketing automation
Airline KLM is using chatbots to drive engagement with customers, helping them get booked on flights in a quicker and more efficient manner.
It started with a simple chatbot on Messenger that was focused on helping customers book travel and avoid having to go through the process on a website. But their strategy evolved, letting the chatbot help people through the entire journey.
KLM is looking into using the chatbot to do more than just help book travel. It also is helping customers with acquiring boarding passes and luggage tags.
Another example Clyde mentioned was a healthcare company using a chatbot to do high-level diagnosis and share real-time physician input about what a person was experiencing.
It used AI to generate responses and give people the best type of recommendation about what might be going on with their health.
Chatbots can also help people schedule for an appointment at urgent care. What this means is that patients can get scheduled for an appearance at urgent care without having to wait in a long line and do it in person.
WeChat is a popular messaging and payment app in China that has nothing to do with websites or emails but still drives customer engagement. People use it for both their personal and work life for it. It lets you collaborate with your coworkers and order an Uber all on the same app.
Cautionary tale: Tay
The chatbot Xiaoice was so popular in China that Microsoft tried to do something similar in the U.S. with a chatbot called Tay.
Tay had to learn how to speak and interact with people, so Microsoft let it loose on Twitter, where it would absorb all sorts of content in order to learn language.
A great idea, right? Not so much. Tay quickly became an inflammatory bot, spouting racist and sexist rhetoric, influenced by the absolute worst parts of Twitter. It was shut down within a day.
4. AI Is Still in Its Early Stages
Most of the bots you’ve interacted with are probably not AI bots. They tend to be scripted.
Someone literally wrote out all the things the bot would say and all the responses it would give based on specific customer questions.
It’s a good start, but it can’t possibly account for the ways a person might ask a question or what kind of problem they might have that needs to be addressed.
AI hasn’t reached sci-fi levels of omnipotence yet. While automation is powerful, the chatbot still very much limited to the imagination and abilities of the creator.