✍️ Naming Your Automations

The first step in crafting an automation is assigning it a unique name. This name should be precise, succinct, and should mirror the action or information that your customers are seeking. Below are some guidelines to help you name your automations effectively:

1. Try to use the exact questions your customers typically ask or begin with a verb such as "Inquire", "Get", or "Request". This strategy aids the AI model in accurately determining the intent of the messages.
    
2. Steer clear of ambiguous automation names as they might lead the AI model to misinterpret the intent of the messages.
   
   Examples:
   ❌ Order info
   ✅ Recommended: Get Order status
   
   ❌ Refund
   ✅ Request for a refund

Formats

The format to be used depends on the scope of the automation that is needed:

1. Inquiry about / Inquiries about
2. Request for
3. Issues about
4. Phrase it as a Question
5. Phrase it as a Statement

1. Inquiry about

This allows you to create an automation that would answer any questions about an Intent. It basically translates to “Any questions about a subject or intent”. This can be effective if we narrow it down to a main “subject” and not make it too vague.

Example automation name: “Inquiry about refunds”

This will be used by Siena as the source of information about anything that is related to Refunds.

However, we do not recommend this for Intents that cover a broad scope, such as “refunds”. When used for “refunds”, it would encompass everything about refunds, including

• Refund policy
• Who are qualified for refunds?
• How long do refunds take?
• Sending orders back for a refund
• Others

Using this format (Inquiry about) tells Siena that she should use the automation for any concern that includes the “word” or “Intent” that we assign to it. A better use for this would be some examples below, as they can have a more focused and targeted response.

• Inquiries about the warranty policy
• Inquiry about allergens
• Inquiry about affiliate programs
• Wholesale inquiry
• Inquiry about shipping locations
• Inquiry about free shipping

💡 Important: When we use this format, Inquiry about”, we need to make sure that we add all the information available about a specific Intent. This will be discussed more when we touch on actual automation responses.

2. Request for

This is used to provide info about an Intent or something that a customer might be asking for.

Example automation name: Request for refund

This translates to an actual request to get a refund, and not a broad question about refunds in general.

Using this format (Request for) tells Siena to provide information about a particular Intent. Unlike the previous format, the response here is designed to give a specific response or answer about a particular subject or intent.

Comparison:

Inquiry about refunds: This can cover everything about refunds, including

• Refund policy
• Who is qualified for refunds
• How long do refunds take
• Sending orders back for a refund
• Others

Request for refund: This focuses mainly on the request of the customer for a refund. We merely need to outline our response on how we want to handle this request.

3. Issues about

This is another “general” automation as it can cover a multitude of topics related to the Intent assigned to it. This is usually used to handle any product issues.

Example automation name: “Issues about broken tumblers”

This will be used by Siena as the source of information about anything that is related to broken tumblers. A common practice is to group common product issues especially if the resolution that will be offered for each issue is similar. Here are scenarios that can have similar resolutions.

Product Issues:

• Broken tumbler
• Cracked lid
• Leaking tumbler

Potential (similar) Resolution:

• Asked for a picture of the damage.
• Offer a partial refund for them to keep the item.
• Offer a replacement.

In cases like this, we can use a forward slash (/) in the automation name to include the different issues with the tumbler.

Automation Name: Issues about damaged tumbler / cracked lid / leaking

4. Phrase it as a Question:

This is the simplest format that you can use, and it simply mimics a customer's intent in the form of a question. Here are the most common examples:

• Is this vegan
• Do you offer free shipping

💡 IMPORTANT: Although these are stated as questions, never place a question mark (?) at the end of the automation name.

Using the other formats would make the automation name too long and seem rather odd.

1. Inquiry about whether the products are vegan
2. Request for information on whether the products are vegan
3. Issues about - this one does not even work for the sample intents.

5. Phrase it as a Statement:

This is the easiest format, as we only need to state the intent as a short phrase or statement.

• Cancel subscription
• Discount did not work
• Account log in issues
• Ingredients of the milkshake
• Order arrived damaged
• Order shows delivered but not received
• Out of stock items/restocking

💡 Pro Tip:

Leverage Expressions: To enhance Siena's ability to accurately classify customer intents, you can incorporate Expressions. These are alternative phrases that a customer might use. For instance, instead of saying "My device is not turning on", some customers might say "My device is not working" or "My device is defective". You can include these in the Expressions section under the automation name.

This strategy enables Siena to comprehend the various ways customers might phrase their requests, thereby improving its ability to identify their intents. At present, you can add up to 5 expressions per automation.

If you have any further questions or need additional assistance, please don't hesitate to reach out to our Customer Support team. We're always here to help!