Going from horse to automobile, the fascination of the Industrial Revolution, and the wow of Artificial Intelligence (AI) have all made humans more able to accelerate a process with less effort. Or at least that was the intent. This can be a double-edged sword. And you know this if you have had to deal with a “stupid bot.” Like artificial sweeteners, too much of what seems a good thing can create new issues. That is where UX can help AI.
AI is the secret sauce of automation in terms of how fast machines can do things we humans would otherwise have to do. And in many cases they can do it much faster and with better accuracy. But have you ever heard of the term garbage in, garbage out? Put simply, the more you accelerate a bad process the more garbage you get out. When this comes to Customer Experience (UX) you can quickly go from pissing off a few customers to pissing off everyone! Humans naturally vary in their behaviors, abilities, and emotional intelligence. A Machine does what it is programmed to do every time. (OK, now with advanced AI models machines can learn. But who are they learning from?)
Using our bot example, if a bot asks you the same question repeatedly can you make the bot understand that you already answered that question? The examples of bad user experiences is a list longer than Santas! So let us stick with the simple case of how to develop a worthwhile bot.
Start with the answer when designing the User Experience your bot will mimic. And the answer is “a smile.” Working backward from a smile takes you all the way back to a frustrated customer who is limited on both time and patience. Each step in the process of using AI to serve this customer efficiently and effectively is a link in the chain leading to success. Break one link and voila, you pissed off another customer.
And for those who believe they need a bot solely because they cannot handle the volume of customers needing assistance, your bot is serving the symptom, not the problem. Do the root cause analysis to find out what is defective in your products and/or services. After that is fixed return to designing an AI-driven process to assist customers with meaningful challenges.
Here is a simple example of what to consider when designing an effective bot.
