Mistakes with AI have been made early on, causing apprehension towards automated chats. Let’s take a closer look at questions you can ask vendors before you commit to a solution.
AI and technology can help make things move faster, but you want to move the right things faster, not the wrong things. Technology can make things more efficient, but there’s also ways where things fail.
The amount of companies who’ve invested or plan to invest in chatbots solutions has jumped by 500% in the last year alone. 80% of users are happy with the technology. – Madeline Laurano, Principal Analyst and Founder of Aptitude Research
Chatbots are by no means representative of all the forms of things that can be helpful, and lately – a hot topic. There’s been areas where they’ve provided value, and there’s been areas where they have clearly failed. When you move beyond the chat and move towards transactions, expectation setting, and using chatbots as schedule versus just talk, there is a tremendous amount of value.
Failure to examine data.
One quick way it can fail is if it’s trained on it on a bad set of data. Ask your vendors, how was, how are the chat-bots making decisions? How are they getting smarter? What data are they using? Is it your data? Is it third party data? Are these chatbots rigid? Are they trying to do too much?
I was talking to a vendor and they said, ‘70% of people that talk to our chatbot think they’re talking to a real human’. While this may sound appealing at first, metrics shouldn’t be tricking people and making them think they’re talking to a person. It’s important, to be honest, and grounded in the truth that this is a piece of technology that can be helpful.
Poor user experience.
There’s a lot of bad end-user experiences. Think back to when you’ve called into the bank or when talking to a chatbot on a website, the truth is people are not using a chatbot to chat, they are here to get things done. If the chat interface is efficient, then people will want to use it. In the talent acquisition context, a chatbot can screen, schedule, and re-engage with candidates, not just chat with you.
Slow turnaround.
The other piece is around the data. Oftentimes we talk to folks that implement AI and they realize that they haven’t thought through the data piece of it, ‘Oh, by the way, it’s going to take six months before this thing actually works because it needs to be trained on our internal data.’ or, ‘Before we implement this AI solution, we have to custom configure it.’ Being able to implement and move quickly to integrate an AI solution is critical.
These are all things to consider when thinking about any form of AI, including chatbots. AI and technology present many exciting opportunities, but they can fail if not used in the right ways. AI replaces the tasks people are doing, but it’s not replacing people. And I think that’s a big difference. By replacing repetitive tasks with true engagement is an incredible opportunity.
