Best practices

Overview

Greetings are your AI Agent’s first impression. They introduce the Agent, set expectations for the conversation, and reflect your brand’s tone. Think of the greeting as the handshake that starts every interaction and guides end users toward what the Agent can do.

A good greeting is clear so end users immediately know who they’re interacting with. It stays focused by outlining exactly what the Agent can do. The tone and style should be on-brand, matching your organization’s personality. And finally, the message must be concise, keeping the conversation moving without overwhelming the end user.

Writing style

Strong greetings are written in a natural, conversational tone. They avoid jargon and overly formal language, and instead use simple sentences that sound like real speech. A well-written greeting feels approachable and invites end users to engage right away.

Many end users are used to simple search or keyword-based chatbots. It’s important to set expectations upfront: Ada works best when people write in full sentences or longer messages. A greeting that encourages this kind of input helps guide end users toward more effective interactions.

Example: Rewrite for clarity

Outcome: The greeting feels human, brand-consistent, and easy for end users to act on while encouraging them to use full sentences for better results.

Compliance messaging

Many enterprise customers use greetings to include compliance-related information. This can help set clear expectations by letting end users know they are interacting with an AI Agent and by linking to relevant company policies (such as privacy, data use, or terms of service).

Guidelines:

• Clearly state that the conversation is with an AI Agent.
• Link to policies (e.g., privacy policy, terms of service, acceptable use) on your brand’s website.
• Keep the compliance message short and pair it with a friendly introduction so the greeting doesn’t feel overly formal or intimidating.

Example: Referencing compliance details

To add compliance messaging to your Greeting:

1. Add a Text block and write your greeting.
2. Add a Link block.
3. In the URL field of the Link block, type the URL of your compliance policy.
4. In the Link Text field, enter the text you want to display for the link.

Personalization and variations

Ada offers multiple ways to add personalization data to your AI Agent. Personalization helps end users feel recognized, while variations prevent conversations from starting the same way every time. Together, these techniques create greetings that feel more engaging and less repetitive.

Example: Personalized greeting

When setting this up, add a fallback value so the greeting still works if the end user’s name isn’t available. For example:

Outcome: End users feel recognized when data is available, and the fallback ensures the greeting remains smooth and natural when it isn’t.

Example: Greeting variations

Outcome: Repeat visitors experience fresh, dynamic greetings.

Temporary or promotional greetings

Greetings can also be updated to share time-sensitive information such as limited-time promotions, special announcements, or recurring events. Adjusting your greeting for these occasions ensures end users see relevant updates immediately, without needing to search elsewhere.

Example: Scheduled greeting

Here are the ways you can implement your Text block within a Greeting together with a Scheduled block:

To set up a scheduled Greeting:

1. Create a schedule for Tuesdays, 3–6 PM.
2. Add a Scheduled block.
3. In the Yes path of the Scheduled block, write the greeting for your promotion.
4. In the No path of the Scheduled block, write the fallback/default greeting that appears outside the promotion window.

Outcome: End users see a promotional greeting during the defined schedule (Tuesdays, 3–6 PM), and a clear alternate greeting at all other times.

Contextual greetings

Not every end user is the same, and greetings should adapt accordingly. Use the Conditional block to implement contextual greetings. The Conditional block allows you to define one or more conditions (for example, user type, region, or language). If none of those conditions are met, the Agent will fall back to the default greeting you’ve configured.

Example: New vs. returning user greeting

To set up a contextual Greeting for new vs. returning end users:

1. Ensure you have a way to determine whether an end user is new or returning. One common approach is to set up a meta variable (e.g., is_returning_user) or use another user attribute passed from your system.
2. Insert a Conditional block.
3. Add a condition for returning end users (using the variable/attribute above) and write an applicable greeting.
4. Add another condition for new end users and write a different greeting.

Outcome: New end users feel guided, while returning end users get a warm welcome.

Example: Region-based greeting

To set up a region-based Greeting:

1. Add a Conditional block, add a condition for the UK region, and write the applicable greeting.
2. Add another condition for the US region and write a different greeting.
3. Add an Else condition to handle all end users outside of the US and UK.

Outcome: End users see localized information that matches their context, while others see your default greeting.

Channel-specific considerations

Not all greetings behave the same way across channels.

• Chat: Keep greetings short, conversational, and aligned with your AI Agent’s persona. Emojis can help make them feel approachable if they fit your brand.
• Voice: Keep greetings short and direct, make it clear the caller is speaking with an AI Agent, and start with a simple question like How can I help you today? (see more in Voice greetings and handoffs).
• Email: Greetings are not supported. Instead, the AI Agent can start replies with Hello [first_name], except during handoffs. To include a greeting in a handoff, add it manually in the first Text block of the flow.

Review and maintenance checklist

Greetings are not set and forget! They need to be reviewed and refreshed regularly. Always preview greetings with and without personalization, test conditional logic to ensure the right greeting is shown, and confirm the tone matches your brand. Update greetings ahead of seasonal events or promotions, and remove outdated content so end users never see irrelevant information.

Recap: Key highlights

The following table summarizes the main scenarios and best practices for configuring greetings, along with the features that support them.