Skip to main content
Join the Agentforce Hackathon on Nov. 18-19 to compete for a $20,000 Grand Prize. Sign up now. Terms apply.

Plan Your Bot Content

Learning Objectives

After completing this unit, you’ll be able to:
  • Explain bot terminology.
  • Describe how to plan for bot content.
  • List some bot content considerations.

Chatbot Terminology

Sita Nagappan-Alvarez, the CEO of Ursa Major Solar, is curious to learn more about what Maria Jimenez, her Salesforce admin, has discovered about chatbots. After Maria mentions that bots are a smart customer channel that requires a lot of content, Sita’s confused. Bot content? She thought you simply click a button or write some code to deploy bots.

Maria reassures Sita that bot deployment is easy. It’s creating a useful, helpful, and relevant bot for customers that’s the challenge. Figuring out the purpose of the bot—and the content it should display—is critical. Planning what the bot does and doesn’t do, and says and doesn’t say, can prevent bad customer experiences.

To understand the types of bot experiences to craft content for, Maria reviews this Einstein bot terminology with the service team.

Terminology  Description
Dialog

Dialogs are conversation snippets that control what a bot can do. Each dialog includes a dialog intent, which is optionally available for NLU training to understand different kinds of customer responses.

During a conversation with a customer, a bot moves between several different dialogs. Each dialog handles a portion of the conversation. For example, Welcome, Main Menu, Order Status, Location and Hours, and Transfer to Agent are individual dialogs that a customer might experience as part of a single conversation with a bot.

A dialog for "Chat with an agent"...might look like this:

Customer: Transfer me to an agent. [Or any variations on this phrase.]
Bot: No problem. Hang on: I’ll connect you with an agent.

Dialog Intent

Dialog intents are the customer's reasons for interacting with your bot. For example, buying a solar panel, returning a solar panel, getting store hours.

Dialog intents are actions and generally start with verbs, such as buy, return, find, look up, cancel, and edit.

Optionally, add dialog intents to your dialogs. Then train the bot to create a learning model that the bot can use to understand dialog intents. If your customers interact with your bot by typing a message in the chat window, use intents to help your bot understand what they want.

Note: If your customers interact with your bot using only menus or buttons, rather than free-form text input, intents aren't needed.

Customer input that relates or triggers an intent may look like this: Where’s my order?

Entity

Entities are a type of data that you want to collect from a customer. Salesforce provides the following system entities: Text, DateTime, Date, Money, Number, Person, Location, Organization, Percent, Boolean, and Object (standard Salesforce or custom). You can create custom entities as needed.

Here are some examples of entities:

  • Order number
  • Email address
Variable A variable is a container that stores a specific piece of data collected from a customer. You must associate each variable with an entity. Since variables are containers of information, you can use them within dialog actions as both inputs and outputs and insert them as text in messages.

Plan Chatbot Content

With an understanding of chatbot terms, Maria and the Ursa Major Solar service team are in a better position to plan what kinds of bot content to create. After all, what good is a bot if it’s not clear what it’s going to say to customers? Not knowing what a bot should say is as bad as the bot saying the wrong thing: Each makes for a poor customer experience.

In her previous planning meeting with the service team, Maria agreed to write the bot’s greeting and compile a list of ways customers ask for help.

Maria explains to Sita and the service team that from everything she’s learned, the best people to research or suggest bot content are the support agents who work with customers every day. Agents know customers. Agents are on the front lines, so they know customers’ questions, concerns, and common issues. But the best people to actually craft bot content are writers. Writers know the importance of words. Writers know how to use words to reflect a company’s brand and tone. Most important, writers know how to create a customer experience with words. Words in, words out… Knowing the impact of words is the best way to write content that helps customers.

Nevertheless, Maria leads the content planning with these topics and questions.

Topic Questions Answers
Define bot context What’s your company name and industry? Ursa Major Solar and the renewable energy industry
What are your most common customer service scenarios? Order status, appointment scheduling, inquiries about prices, store hours, and locations
Give the bot personality What’s your bot’s name? Solar Sammy
Describe your bot’s personality using one to three adjectives. Smart, snappy, and sincere
How is this similar to or different from your company brand? Similar to our brand, but don’t let smart become arrogant
How does your bot convey this personality in a greeting? For example, hello, HELLO, Hello!, Hi, Hey, ‘Sup, how can I help u? “Hello, I’m a bot named Solar Sammy!”
How does it convey this personality in a farewell? “Goodbye, I hope I helped you.”
How does it respond to gratitude with personality? “Oh, thank you!”
How does it apologize with personality? “Oh, I’m sorry. I’d like to help you.”
When should your bot change the personality of its apologies? If Solar Sammy makes more than two apologies, it should redirect to a support agent immediately.
Design the conversation Can you identify the bot as not human in a greeting? Yes: “Hello, I’m a bot named Solar Sammy! I’m here to help, but a real human can help you at any time, too.”
Are there any menu options to set expectations? Yes: We want the option to transfer to an agent at any time, and to close the conversation, to show in a persistent menu.

Chatbot Content Considerations

Along with planning content for chatbot conversations, Maria and the service team also discuss these content considerations. As they all know by now, the best way to avoid bad customer experiences with bots is to plan ahead.

  • Openings: Opening a chat with a question—like "How can I help you?"—gets down to business right away.
  • Closings: "Thank you" can end a chat. Customers may indicate they're ready to leave the conversation with words like "OK," "Well," and "All right."
  • Response Delays: Super-short pauses—say, less than one second—between conversation snippets can seem artificial. The average human pause time in text chats is approximately two to four seconds.
  • Emojis and Emoticons: Emojis are warm and enthusiastic: 😖 😄. Emoticons are lukewarm and mildly enthusiastic (e.g, :) ^_^ >_>). You can use emojis in place of whole words (👍 = "OK," "good").
  • Text Style: Machine-like fonts, such as Roboto Mono and Orbitron, clarify to customers that they're interacting with a bot. ALL CAPS equals yelling, all lowercase is informal, and utterance-final period (for example, "I'm fine.") can equal seriousness, formality, distance, irritation, or sarcasm.
Note
Looking for sample content or apps to extend your bot? Then check out AppExchange! Salesforce has many partners who use the flexibility of the Salesforce Platform to build amazing solutions to extend Salesforce easily.

With a better understanding of what makes helpful bot content, Maria and the service team are ready to take the next step of learning about Einstein bots.
Share your Trailhead feedback over on Salesforce Help.

We'd love to hear about your experience with Trailhead - you can now access the new feedback form anytime from the Salesforce Help site.

Learn More Continue to Share Feedback