Have you heard about “the one thing” productivity hack? If not, it’s the ONE thing you can do such that by doing it everything else will be easier or unnecessary.
Chatbots too, have ONE thing.
I’ll get to that in a minute but first, some backstory. We built a natural language chatbot using IBM Watson. IBM has a solid suite of tools, cognitive services, and cloud infrastructure required to power sophisticated AI-powered applications.
A chatbot is one such application. As we got familiar with the Watson environment, capabilities, and limitations we came to places where we debated going left, right, or straight ahead.
If you’re thinking about deploying an NLU chatbot I have a set of guidelines learned from doing it.
5 Strategic Recommendations
1. Have a Vision. Because vision is the ONE thing. Who is the bot going to help, in which situations, and how is it going to help? Before we started to make our bot I envisioned a person driving through town, both hands on the steering wheel, asking Siri to book a flight. You can read how we got to that point in AI Will change the customer experience: a look five years forward.
2. Build to your vision. At one point during development it became clear that dialog buttons could be used to expedite the project. Dialog buttons are Yes/No; Small/Med/Large, options in a conversation and are popular in many chat UIs.
When we hit that point in development the thought occurred “should we or shouldn’t we?” That’s where our vision led the way. Can a motorist with both hands on the wheel engage in chatbot dialog with buttons to press? There would be no buttons.
3. Keep it simple. Natural language UIs need to be trained in the language of the domain in which they are used. The tighter the bot’s purpose, the less data it needs for training. Master a select few tasks for easier development and ongoing improvement. In the near term there’s no such thing as a general AI chatbot. Focus on and master one or two tasks that make customers happy and achieve a business goal.
4. Understand the data flow. Chatbot conversation can take place through any number of UIs ranging from Facebook Messenger to smartphone assistants to your own app. You’ll want to understand who has access to your data, how it’s governed by your own and regulatory policies, and how it is secured.
5. Analyze. Natural language dialog is a new and rich window into the customer experience. Your pilot program should include the ability to capture and analyze this information to improve bit performance.