Setting Your Bot's Personality & Tone

Setting Your Bot's Personality & Tone

Your bot's personality is how it "sounds" when it talks to visitors — warm and casual, or polished and formal. This page walks through the Persona tab, where you shape that voice. Everything here changes the words your bot chooses, not what it knows.

Where to find it

1. Go to Chatbots in the left sidebar and click your bot.
2. Click Configure in the row of tabs.
3. In the left menu, click Persona.

Every setting below has a live example right on the page, so you can see the difference before saving.

Conversation tone — the big one

What it is: the overall style of the bot's replies. Why it matters: it's the first thing a visitor notices. How: pick one of three cards.

• Professional — formal and precise. "Thank you for reaching out. I would be happy to assist with your invoice inquiry."
• Friendly — warm and approachable, with the occasional emoji. "Of course, I'd love to help! 😊 Could you share your invoice number?"
• Casual — relaxed and fun. "Sure thing! Drop your invoice number and I'll take a look 👀"

After you pick one, a "How it sounds" example appears showing a sample back-and-forth in that tone.

Bot mood — the vibe underneath the words

What it is: the deeper personality flavor layered on top of tone. Why it matters: two "Friendly" bots can still feel different — one bubbly, one soothing. How: pick a mood card.

• Friendly — warm & personable
• Professional — polished & precise
• Energetic — upbeat & dynamic (great for launches)
• Calm — serene & reassuring (great for support/healthcare)
• Formal — proper & courteous
• Sales-Driven — persuasive & value-focused (nudges toward buying)

Example: set mood to Sales-Driven and the bot naturally mentions relevant upgrades; set it to Calm and it slows down and reassures.

Empathy style — how it handles feelings

What it is: how the bot reacts when a visitor is upset, confused, or excited. Why it matters: the right emotional response keeps people from getting frustrated. How: pick one.

• Empathetic — validates feelings first ("I completely understand how frustrating that is…").
• Assertive — confident, gets straight to the fix.
• Neutral — balanced and fact-focused.

Intent sensitivity — how eagerly it acts

What it is: how quickly the bot decides a visitor wants something done (like escalating to a human). Why it matters: too eager and it hands off constantly; too cautious and it misses real requests. How: choose Low (only obvious requests), Medium (balanced — a good default), or High (proactive, catches subtle hints).

Sentiment response rules — auto-adjust on the fly

What it is: on/off rules that change behavior based on the visitor's mood during a chat. Why it matters: it lets one bot handle an angry customer differently from a happy one, automatically. How: flip the toggles you want:

• Negative sentiment → increase empathy
• Angry customer → apologize and offer a human
• Confused customer → switch to step-by-step explanations
• Positive sentiment → mention relevant upgrades

Response length — short or thorough

What it is: how much the bot writes per answer. Why it matters: long answers annoy people in a chat window; too short can feel unhelpful. How: pick one.

• Concise — 1–2 sentences, direct.
• Balanced — moderate detail (a safe default).
• Detailed — thorough explanations.

Welcome message — the greeting

What it is: the first thing a visitor sees when they open the chat. Why it matters: a good opener invites people to actually type something. How: in the Messages section, edit the Welcome Message box. Example: Hi! How can I help you today?

Fallback message — when it's stumped

What it is: what the bot says when it doesn't understand or doesn't know. Why it matters: a graceful "I'm not sure, let me connect you" is far better than a confusing wrong answer. How: edit the Fallback Message box. Example: I'm not sure about that — let me connect you with our team.

System instructions — the rulebook you write

What it is: a free-text box where you tell the bot how to behave, in your own words. These instructions are quietly added to every conversation. Why it matters: it's the most powerful control here — you can set your brand voice, off-limits topics, and escalation rules all at once. How: in System instructions, write plain-English rules. Example:

You are a helpful support agent for Acme Corp. Always be polite, stay on topic, and escalate billing issues to a human agent.

Be specific about your brand voice, what the bot should not do, and when to hand off.

Save and check

Save your changes, then open the Test Bot preview (top-right of the bot's dashboard) and ask a few questions. If the voice isn't quite right, tweak the tone or system instructions and test again. See Testing Your Bot Before You Launch.

Related

• Customizing How Your Widget Looks
• Guardrails: Confidence, Limits & Requiring an Email
• Training Your Chatbot